Sync main snapshot to Hugging Face (no local binary banner)

.gitattributes

@@ -1,8 +1,7 @@
-* text=auto eol=lf
+* text=auto eol=lf
 *.md text eol=lf
 *.py text eol=lf
 *.tsv text eol=lf
-
 *.7z filter=lfs diff=lfs merge=lfs -text
 *.arrow filter=lfs diff=lfs merge=lfs -text
 *.bin filter=lfs diff=lfs merge=lfs -text
@@ -38,3 +37,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+

.github/workflows/resource_sync.yml

@@ -32,6 +32,36 @@ jobs:
       - name: Validate catalog
         run: python scripts/validate_resource_catalog.py
 
+      - name: Ensure labels exist
+        uses: actions/github-script@v7
+        with:
+          script: |
+            const labels = [
+              { name: "resource-update", color: "0e8a16", description: "Automated resource catalog updates" },
+              { name: "needs-review", color: "fbca04", description: "Needs maintainer review before merge" }
+            ];
+            for (const label of labels) {
+              try {
+                await github.rest.issues.getLabel({
+                  owner: context.repo.owner,
+                  repo: context.repo.repo,
+                  name: label.name
+                });
+              } catch (error) {
+                if (error.status === 404) {
+                  await github.rest.issues.createLabel({
+                    owner: context.repo.owner,
+                    repo: context.repo.repo,
+                    name: label.name,
+                    color: label.color,
+                    description: label.description
+                  });
+                } else {
+                  throw error;
+                }
+              }
+            }
+
       - name: Create review PR
         uses: peter-evans/create-pull-request@v6
         with:

CITATION.cff

@@ -0,0 +1,21 @@
+cff-version: 1.2.0
+message: "If you use this repository, please cite it."
+title: "Pashto Language Resources Hub (Pukhto/Pashto)"
+type: software
+version: 0.1.0
+license: Apache-2.0
+repository-code: "https://github.com/Musawer1214/pashto-language-resources"
+url: "https://github.com/Musawer1214/pashto-language-resources"
+authors:
+  - family-names: "Musawer"
+    given-names: "Musawer"
+keywords:
+  - Pashto
+  - Pukhto
+  - Pushto
+  - ASR
+  - TTS
+  - NLP
+  - machine translation
+  - language resources
+

README.md

@@ -1,63 +1,57 @@
----
-license: apache-2.0
-language:
-- ps
-tags:
-- pashto
-- asr
-- tts
-- nlp
----
+# Pashto Language Resources Hub (Pukhto/Pashto)
 
-![Pukhto Pashto Repository Banner](Repository_banner_Image.png)
+Open-source repository for Pashto language technology resources: datasets, models, benchmarks, ASR, TTS, NLP, and machine translation (MT).
 
-# Pukhto/Pashto Open Language Project
+This project curates verified Pashto resources and maintains reproducible tooling for discovery, validation, and documentation.
 
-Community-led open-source project to make Pashto a first-class language in speech and language technology.
+![Pukhto Pashto Repository Banner](https://raw.githubusercontent.com/Musawer1214/pashto-language-resources/main/Repository_banner_Image.png)
 
-## Project Links
-- GitHub Pages (About): [Pukhto_Pashto Site](https://musawer1214.github.io/Pukhto_Pashto/)
-- GitHub Pages (Resource Search): [Pashto Resource Search](https://musawer1214.github.io/Pukhto_Pashto/search/)
+## Start Here
+
+- Main resource search: [Pashto Resource Search](https://musawer1214.github.io/pashto-language-resources/search/)
+- Project site: [Pashto Language Resources Hub](https://musawer1214.github.io/pashto-language-resources/)
+- GitHub repository: [Musawer1214/pashto-language-resources](https://github.com/Musawer1214/pashto-language-resources)
+- Hugging Face mirror: [Musawer14/pashto-language-resources](https://huggingface.co/Musawer14/pashto-language-resources)
+
+## If You Searched For
+
+This repository is relevant to these search intents:
+
+- Pashto datasets
+- Pashto ASR model
+- Pashto TTS resources
+- Pashto NLP benchmark
+- Pashto machine translation resources
+- Pukhto language technology
+- Pushto AI resources
 
 ## Current Scope
+
 - Build open Pashto datasets, benchmarks, and model references for ASR, TTS, NLP, and MT.
-- Track practical tools, apps, and academic papers relevant to Pashto integration in technology.
+- Track practical tools, apps, and academic papers for Pashto integration in technology.
 - Keep everything transparent, reproducible, and contribution-friendly.
 
-## Resource System (Current)
+## Resource System
 
-This repository now has a machine-readable and searchable resource pipeline:
+Machine-readable and searchable resource pipeline:
 
 - Canonical catalog: [resources/catalog/resources.json](resources/catalog/resources.json)
 - Catalog schema: [resources/schema/resource.schema.json](resources/schema/resource.schema.json)
 - Candidate feed (auto-generated): [resources/catalog/pending_candidates.json](resources/catalog/pending_candidates.json)
-- Search UI: [docs/search/index.html](docs/search/index.html)
+- Search UI source: [docs/search/index.html](docs/search/index.html)
 - Search data export: [docs/search/resources.json](docs/search/resources.json)
-- Full index docs: [docs/resource_catalog.md](docs/resource_catalog.md)
+- Resource index docs: [docs/resource_catalog.md](docs/resource_catalog.md)
 - Automation docs: [docs/resource_automation.md](docs/resource_automation.md)
-- Repeatable runbook: [docs/resource_cycle_runbook.md](docs/resource_cycle_runbook.md)
+- Cycle runbook: [docs/resource_cycle_runbook.md](docs/resource_cycle_runbook.md)
 
 ## How New Resources Are Added
 
-The process is semi-automatic:
-
-1. Auto discovery:
-- Daily GitHub Action runs `.github/workflows/resource_sync.yml`.
-- It updates `resources/catalog/pending_candidates.json` and opens a review PR.
-
-2. Manual review and promotion:
-- Maintainers inspect candidate quality, Pashto evidence, and license/usage compatibility.
-- Approved entries are moved into `resources/catalog/resources.json` with `status: verified`.
-
-3. Regeneration and validation:
-- Run `python scripts/validate_resource_catalog.py`
-- Run `python scripts/generate_resource_views.py`
-- Commit generated updates (`resources/*/README.md` and `docs/search/resources.json`).
+1. Auto discovery runs daily from `.github/workflows/resource_sync.yml` and updates `resources/catalog/pending_candidates.json` in a review PR.
+2. Manual review checks quality, Pashto evidence, and license compatibility before promoting entries into `resources/catalog/resources.json` with `status: verified`.
+3. Regeneration and validation runs `python scripts/validate_resource_catalog.py` and `python scripts/generate_resource_views.py`, then commits generated updates.
 
 Shortcut wrapper:
-- Run `python scripts/run_resource_cycle.py --limit 25`
-
-This prevents low-confidence links from being merged directly while still automating discovery.
+- `python scripts/run_resource_cycle.py --limit 25`
 
 ## Quickstart
 
@@ -69,7 +63,15 @@ python scripts/check_links.py
 python -m pytest -q
 ```
 
+## Discoverability And SEO
+
+- Playbook: [docs/discoverability_seo.md](docs/discoverability_seo.md)
+- Docs hub: [docs/README.md](docs/README.md)
+- Resource search page: [docs/search/index.html](docs/search/index.html)
+- Citation metadata: [CITATION.cff](CITATION.cff)
+
 ## Documentation Map
+
 - Purpose: [PROJECT_PURPOSE.md](PROJECT_PURPOSE.md)
 - Contributing: [CONTRIBUTING.md](CONTRIBUTING.md)
 - Roadmap: [ROADMAP.md](ROADMAP.md)
@@ -82,6 +84,7 @@ python -m pytest -q
 - Resource automation: [docs/resource_automation.md](docs/resource_automation.md)
 
 ## Resource Sections
+
 - Datasets: [resources/datasets/README.md](resources/datasets/README.md)
 - Models: [resources/models/README.md](resources/models/README.md)
 - Benchmarks: [resources/benchmarks/README.md](resources/benchmarks/README.md)
@@ -91,6 +94,7 @@ python -m pytest -q
 - Code: [resources/codes/README.md](resources/codes/README.md)
 
 ## Workspaces
+
 - [data/](data/README.md): datasets, curation, metadata, quality
 - [asr/](asr/README.md): ASR baselines and experiments
 - [tts/](tts/README.md): TTS baselines and experiments
@@ -98,3 +102,6 @@ python -m pytest -q
 - [experiments/](experiments/README.md): reproducible run cards
 - [apps/desktop/](apps/desktop/README.md): user-facing integration references
 - [models/](models/README.md): model layout and release conventions
+
+
+

benchmarks/results/templates/tts_result.example.json

@@ -4,7 +4,7 @@
     "name": "internal_prompt_set",
     "version": "v1",
     "split": "eval",
-    "source_url": "https://github.com/Musawer1214/Pukhto_Pashto"
+    "source_url": "https://github.com/Musawer1214/pashto-language-resources"
   },
   "model": {
     "name": "facebook/mms-tts",
@@ -26,3 +26,4 @@
     "config_ref": "tts/configs/mms_tts_eval.yaml"
   }
 }
+

benchmarks/schema/benchmark_result.schema.json

@@ -1,6 +1,6 @@
 {
   "$schema": "https://json-schema.org/draft/2020-12/schema",
-  "$id": "https://github.com/Musawer1214/Pukhto_Pashto/benchmarks/schema/benchmark_result.schema.json",
+  "$id": "https://github.com/Musawer1214/pashto-language-resources/benchmarks/schema/benchmark_result.schema.json",
   "title": "Pashto Benchmark Result",
   "type": "object",
   "required": [
@@ -67,3 +67,4 @@
   },
   "additionalProperties": false
 }
+

docs/README.md

@@ -15,6 +15,7 @@ This folder is the main documentation entry point for contributors.
 - Dataset guidelines: [dataset_guidelines.md](dataset_guidelines.md)
 - Pashto normalization policy: [pashto_normalization_v0.1.md](pashto_normalization_v0.1.md)
 - Common Voice Pashto integration: [common_voice_pashto_24.md](common_voice_pashto_24.md)
+- Discoverability and SEO playbook: [discoverability_seo.md](discoverability_seo.md)
 - Release process: [release_process.md](release_process.md)
 - Release checklist: [release_checklist.md](release_checklist.md)
 - Platforms and publish flow: [platforms.md](platforms.md)

docs/_config.yml

@@ -1,4 +1,21 @@
-title: Pukhto Pashto
-description: Community-led open Pashto AI language project
+title: Pashto Language Resources Hub
+description: Open Pashto (Pukhto/Pashto) datasets, models, ASR, TTS, NLP, MT, and benchmark resources.
 theme: minima
 markdown: kramdown
+url: "https://musawer1214.github.io"
+baseurl: "/pashto-language-resources"
+lang: en
+locale: en_US
+plugins:
+  - jekyll-seo-tag
+  - jekyll-sitemap
+
+logo: "https://raw.githubusercontent.com/Musawer1214/pashto-language-resources/main/Repository_banner_Image.png"
+
+social:
+  name: Pashto Language Resources Hub
+  links:
+    - https://github.com/Musawer1214/pashto-language-resources
+    - https://huggingface.co/Musawer14/pashto-language-resources
+
+

docs/discoverability_seo.md

@@ -0,0 +1,89 @@
+# Discoverability and SEO Playbook
+
+This playbook focuses on making the repository easier to find in:
+
+- GitHub search
+- Google/Bing search
+- Academic and resource discovery channels
+
+## 1) Repository Rename Recommendation
+
+Current slug is `pashto-language-resources`.
+
+Previous slug was `Pukhto_Pashto`, which was less search-friendly due to underscore and mixed spelling.
+
+Recommended slug options:
+
+1. `pashto-language-resources`
+2. `pashto-ai-resources`
+3. `pashto-language-tech`
+
+Selection rule:
+- Prefer the name that starts with `pashto` and includes a clear intent word like `resources`.
+
+After rename:
+- Update `docs/_config.yml` `baseurl`.
+- Update hardcoded URLs in `README.md`, `docs/index.md`, and `docs/search/index.html`.
+- Keep old links alive via GitHub redirect behavior, but still update links in-repo.
+
+## 2) GitHub About Section (Manual UI)
+
+Set these in repository `Settings -> General` and About panel:
+
+- Description:
+  - `Open-source Pashto (Pukhto/Pashto) datasets, ASR, TTS, NLP, MT, models, and benchmark resources.`
+- Website:
+  - [GitHub Pages home](https://musawer1214.github.io/pashto-language-resources/) (or new slug after rename)
+- Topics:
+  - `pashto`
+  - `pukhto`
+  - `pushto`
+  - `asr`
+  - `tts`
+  - `nlp`
+  - `machine-translation`
+  - `speech-recognition`
+  - `language-resources`
+  - `low-resource-languages`
+
+## 3) Content Signals
+
+- Keep the first 160 characters of `README.md` keyword clear.
+- Use consistent terminology across pages: `Pashto (Pukhto/Pushto)`.
+- Publish regular updates in `CHANGELOG.md` and GitHub Releases.
+- Keep `CITATION.cff` updated for scholarly reuse and citation.
+
+## 4) Pages SEO and Crawlability
+
+Already included in this repository:
+
+- `docs/_config.yml` with sitemap and SEO plugin support.
+- `docs/robots.txt` with sitemap reference.
+- Page-level metadata and structured data in `docs/search/index.html`.
+
+Keep these updated when renaming slug or domain.
+
+## 5) External Discovery Boost
+
+- Add the GitHub Pages search URL to:
+  - Hugging Face model/dataset cards
+  - Relevant community profiles and README links
+  - Conference/demo pages for Pashto language technology
+- Ask contributors to link specific resource pages in blog posts or papers.
+
+## 6) Indexing Checklist (After Push)
+
+1. Push all changes to `main`.
+2. Verify GitHub Pages is serving:
+   - `/`
+   - `/search/`
+   - `/robots.txt`
+   - `/sitemap.xml`
+3. Add site property in Google Search Console.
+4. Submit sitemap URL:
+   - [Sitemap file](https://musawer1214.github.io/pashto-language-resources/sitemap.xml)
+5. Run URL Inspection and request indexing for:
+   - Home page
+   - Search page
+6. Recheck search visibility after 1 to 3 weeks.
+

docs/github_operations.md

@@ -1,21 +1,32 @@
-# 🧰 GitHub Operations Guide
+# GitHub Operations Guide
 
-This guide covers repository operations that can be versioned in code and
-manual steps that must be done in the GitHub web UI.
+This guide covers repository operations that can be versioned in code and manual steps that must be done in the GitHub web UI.
+
+## In-Repo Operations (Versioned)
 
-## In-Repo Operations (already versioned)
 - Issue templates: [../.github/ISSUE_TEMPLATE/](../.github/ISSUE_TEMPLATE/)
 - PR template: [../.github/PULL_REQUEST_TEMPLATE.md](../.github/PULL_REQUEST_TEMPLATE.md)
 - CI workflow: [../.github/workflows/ci.yml](../.github/workflows/ci.yml)
 - Release template: [../.github/release_template.md](../.github/release_template.md)
 
-## Manual GitHub UI Steps (not stored in repo)
-- Create/update GitHub Project board.
+## Manual GitHub UI Steps (Not Stored in Repo)
+
+- Create or update GitHub Project board.
 - Define milestone dates.
 - Configure branch protection rules.
 - Configure required status checks.
+- Update repository About description, website, and topics.
+- Upload social preview image.
+
+## Discoverability-Specific GitHub UI Tasks
+
+- About description should include: `Pashto`, `ASR`, `TTS`, `NLP`, and `machine translation`.
+- Topics should include at least: `pashto`, `pukhto`, `pushto`, `asr`, `tts`, `nlp`, `language-resources`.
+- Website should point to GitHub Pages search or docs home.
+- After renaming the repository, update pinned repositories and profile links.
 
 ## Recommended Project Board Columns
+
 1. Backlog
 2. Ready
 3. In Progress

docs/index.md

@@ -1,16 +1,17 @@
 ---
 layout: default
-title: About Pukhto Pashto
+title: Pashto Language Resources Hub
+description: Open-source Pashto (Pukhto/Pashto) datasets, models, benchmarks, ASR, TTS, NLP, and MT resources.
 ---
 
-# About This Repository
+# Pashto Language Resources Hub
 
-`Pukhto_Pashto` is a community-led open project focused on making Pashto a first-class language in speech and language AI.
+`pashto-language-resources` is a community-led open-source project focused on making Pashto (also written as Pukhto/Pushto) a first-class language in speech and language AI.
 
 ## Mission
 
-- Build open Pashto datasets for ASR, TTS, and NLP.
-- Publish reproducible baseline models and evaluation workflows.
+- Build open Pashto datasets and resource indexes for ASR, TTS, NLP, and MT.
+- Publish reproducible baseline models, benchmark schemas, and evaluation workflows.
 - Keep progress transparent, contributor-friendly, and public-benefit oriented.
 
 ## What Is In This Repository
@@ -23,28 +24,41 @@ title: About Pukhto Pashto
 - `docs/`: policies, roadmap, release process, and operating guides.
 - `resources/`: verified external Pashto datasets, models, tools, benchmarks, and papers.
 
-## Search Resources
+## Search Pashto Resources
 
 - Search UI: [Pashto Resource Search](search/)
 - Resource index docs: [resource_catalog.md](resource_catalog.md)
-- Machine-readable catalog: [../resources/catalog/resources.json](../resources/catalog/resources.json)
+- Machine-readable catalog (GitHub): [resources.json source](https://github.com/Musawer1214/pashto-language-resources/blob/main/resources/catalog/resources.json)
 
 ## Project References
 
-- Repository: [Musawer1214/Pukhto_Pashto](https://github.com/Musawer1214/Pukhto_Pashto)
-- Hugging Face: [Musawer14/Pukhto_Pashto](https://huggingface.co/Musawer14/Pukhto_Pashto)
+- Repository: [Musawer1214/pashto-language-resources](https://github.com/Musawer1214/pashto-language-resources)
+- Hugging Face: [Musawer14/pashto-language-resources](https://huggingface.co/Musawer14/pashto-language-resources)
 - Purpose: [PROJECT_PURPOSE.md](../PROJECT_PURPOSE.md)
 - Roadmap: [ROADMAP.md](../ROADMAP.md)
 - Contributing: [CONTRIBUTING.md](../CONTRIBUTING.md)
 
 ## Contributing
 
-You can help by improving documentation, validating normalization rows, sharing verified resources, or contributing data and evaluation workflows.
+You can help by improving documentation, validating normalization rows, sharing verified resources, or contributing data, model, and evaluation workflows.
 
 For contributor workflow and standards, start at:
 - [docs/README.md](README.md)
 - [community/COMMUNICATION.md](../community/COMMUNICATION.md)
 
+## Search Terms
+
+This project is relevant to searches like:
+
+- Pashto datasets
+- Pashto ASR model
+- Pashto TTS resources
+- Pashto NLP benchmark
+- Pashto language technology
+- Pukhto language resources
+
 ## License
 
 This project is released under Apache 2.0. See [LICENSE](../LICENSE).
+
+

docs/platforms.md

@@ -1,27 +1,39 @@
-# 🌐 Platforms
+# Platforms
 
-## 🧭 Primary Platforms
-- GitHub: code, issues, pull requests, releases
-- Hugging Face Hub: models, datasets, demos
-- Community chat (Discord/Matrix): contributor coordination
+## Primary Platforms
+
+- GitHub: code, issues, pull requests, releases, and docs source.
+- Hugging Face Hub: models, datasets, and demos.
+- Community chat (Discord/Matrix): contributor coordination.
+
+## Resource Discovery and Validation
 
-## 📚 Resource Discovery and Validation
 - Use [docs/resource_catalog.md](resource_catalog.md) as the single source of truth for validated external resources.
 - Add new links only after checking official pages and explicit Pashto support markers.
 
-## 📣 Publishing Expectations
-- Every release links to changelog + benchmark snapshot.
-- Every model links to dataset provenance and eval metrics.
-- Every new external link must include use-case notes and where it belongs in repo structure.
-- CI should pass before merging (`.github/workflows/ci.yml`).
+## Publishing Expectations
+
+- Every release links to changelog and benchmark snapshot.
+- Every model links to dataset provenance and evaluation metrics.
+- Every new external link includes use-case notes and target location in repo structure.
+- CI must pass before merging (`.github/workflows/ci.yml`).
+
+## Dual Publish Checklist (GitHub and Hugging Face)
 
-## 🚀 Dual Publish Checklist (GitHub + Hugging Face)
 1. `git status` is clean except intended changes.
-2. Docs and resource links updated.
-3. Commit created with clear scope.
+2. Docs and resource links are updated.
+3. Commit message is scoped and explicit.
 4. Push to `origin` (GitHub).
 5. Push to `hf` (Hugging Face).
 6. Verify README render and link health on both platforms.
 
-## 🧰 Operations Guide
+## Discoverability Checklist
+
+- Keep GitHub About description, topics, and website URL updated.
+- Keep [docs/discoverability_seo.md](discoverability_seo.md) current with slug and sitemap URLs.
+- Ensure links from Hugging Face cards point to both the repository and search page.
+
+## Operations Guide
+
 - GitHub operations and manual UI tasks: [github_operations.md](github_operations.md)
+

docs/resource_automation.md

@@ -13,7 +13,11 @@ This repository uses a semi-automated process to keep Pashto resources current w
 - Hugging Face models
 - Hugging Face Spaces (projects)
 - GitHub repositories (projects and code)
-- Research-paper endpoints
+- GitLab repositories (projects and code)
+- Zenodo records
+- Dataverse datasets
+- DataCite DOI records
+- Research-paper endpoints (arXiv, Semantic Scholar, OpenAlex, Crossref)
 
 ## Files involved
 - Canonical verified catalog: [../resources/catalog/resources.json](../resources/catalog/resources.json)

docs/resource_cycle_runbook.md

@@ -21,7 +21,7 @@ What it executes:
 4. `python scripts/check_links.py`
 5. `python -m pytest -q`
 
-Candidate sources in the sync step include Kaggle datasets, Hugging Face datasets/models/spaces, GitHub repositories, and paper endpoints.
+Candidate sources in the sync step include Kaggle datasets, Hugging Face datasets/models/spaces, GitHub repositories, GitLab repositories, Zenodo records, Dataverse datasets, DataCite DOI records, and paper endpoints (arXiv, Semantic Scholar, OpenAlex, Crossref).
 
 ## Discovery-only mode
 If you only want fresh candidates:

docs/robots.txt

@@ -0,0 +1,5 @@
+User-agent: *
+Allow: /
+
+Sitemap: https://musawer1214.github.io/pashto-language-resources/sitemap.xml
+

docs/search/index.html

@@ -1,9 +1,46 @@
+---
+layout: null
+title: Pashto Resource Search
+description: Search verified Pashto (Pukhto/Pashto) datasets, models, tools, benchmarks, papers, and projects.
+permalink: /search/
+---
 <!doctype html>
 <html lang="en">
 <head>
   <meta charset="utf-8">
   <meta name="viewport" content="width=device-width, initial-scale=1">
   <title>Pashto Resource Search</title>
+  <meta name="description" content="Search verified Pashto (Pukhto/Pashto) resources for ASR, TTS, NLP, machine translation, tools, datasets, models, and benchmarks.">
+  <meta name="keywords" content="Pashto resources, Pukhto resources, Pashto datasets, Pashto ASR, Pashto TTS, Pashto NLP, Pashto machine translation, Pashto benchmarks">
+  <meta name="robots" content="index,follow,max-image-preview:large">
+  <link rel="canonical" href="https://musawer1214.github.io/pashto-language-resources/search/">
+  <meta property="og:type" content="website">
+  <meta property="og:site_name" content="Pashto Language Resources Hub">
+  <meta property="og:title" content="Pashto Resource Search">
+  <meta property="og:description" content="Discover verified Pashto datasets, models, tools, benchmarks, projects, and papers.">
+  <meta property="og:url" content="https://musawer1214.github.io/pashto-language-resources/search/">
+  <meta property="og:image" content="https://raw.githubusercontent.com/Musawer1214/pashto-language-resources/main/Repository_banner_Image.png">
+  <meta name="twitter:card" content="summary">
+  <meta name="twitter:title" content="Pashto Resource Search">
+  <meta name="twitter:description" content="Search and filter verified Pashto language technology resources.">
+  <meta name="twitter:image" content="https://raw.githubusercontent.com/Musawer1214/pashto-language-resources/main/Repository_banner_Image.png">
+  <script type="application/ld+json">
+    {
+      "@context": "https://schema.org",
+      "@type": "CollectionPage",
+      "name": "Pashto Resource Search",
+      "url": "https://musawer1214.github.io/pashto-language-resources/search/",
+      "description": "Search verified and candidate Pashto resources for ASR, TTS, NLP, MT, tools, datasets, and benchmarks.",
+      "inLanguage": "en",
+      "about": [
+        "Pashto datasets",
+        "Pashto ASR",
+        "Pashto TTS",
+        "Pashto NLP",
+        "Pashto machine translation"
+      ]
+    }
+  </script>
   <link rel="preconnect" href="https://fonts.googleapis.com">
   <link rel="preconnect" href="https://fonts.gstatic.com" crossorigin>
   <link href="https://fonts.googleapis.com/css2?family=IBM+Plex+Sans+Arabic:wght@400;500;700&family=Space+Grotesk:wght@500;700&display=swap" rel="stylesheet">
@@ -131,6 +168,26 @@
       font-weight: 600;
     }
 
+    .crawl {
+      margin: 8px 2px 16px;
+      color: var(--muted);
+      font-size: 14px;
+      line-height: 1.5;
+    }
+
+    .crawl p {
+      margin: 6px 0;
+    }
+
+    .crawl-links {
+      margin: 8px 0 0;
+      padding-left: 18px;
+    }
+
+    .crawl-links li {
+      margin: 3px 0;
+    }
+
     .grid {
       display: grid;
       grid-template-columns: repeat(auto-fill, minmax(260px, 1fr));
@@ -273,6 +330,26 @@
       <span class="badge" id="generatedAt">Catalog timestamp: -</span>
     </div>
 
+    <section class="crawl" aria-label="Pashto resource categories">
+      <p>
+        Browse or search resource entries covering Pashto datasets, speech recognition (ASR), text-to-speech (TTS), NLP, translation, tools, and benchmarks.
+      </p>
+      <p>
+        Project overview: <a href="https://github.com/Musawer1214/pashto-language-resources">Musawer1214/pashto-language-resources</a>
+      </p>
+      <ul class="crawl-links">
+        <li><a href="https://github.com/Musawer1214/pashto-language-resources/blob/main/resources/datasets/README.md">Pashto datasets</a></li>
+        <li><a href="https://github.com/Musawer1214/pashto-language-resources/blob/main/resources/models/README.md">Pashto models</a></li>
+        <li><a href="https://github.com/Musawer1214/pashto-language-resources/blob/main/resources/benchmarks/README.md">Pashto benchmarks</a></li>
+        <li><a href="https://github.com/Musawer1214/pashto-language-resources/blob/main/resources/tools/README.md">Pashto tools</a></li>
+        <li><a href="https://github.com/Musawer1214/pashto-language-resources/blob/main/resources/papers/README.md">Pashto papers</a></li>
+      </ul>
+    </section>
+
+    <noscript>
+      JavaScript is needed for filtering, but the linked category pages above remain accessible.
+    </noscript>
+
     <ul id="results" class="grid"></ul>
   </main>
 
@@ -416,3 +493,4 @@
   </script>
 </body>
 </html>
+

docs/search/resources.json

@@ -1,6 +1,6 @@
 {
-  "generated_on": "2026-02-16T00:00:00Z",
-  "count": 75,
+  "generated_on": "2026-02-17T00:00:00Z",
+  "count": 95,
   "resources": [
     {
       "id": "dataset-common-voice-ps-v24",
@@ -1846,6 +1846,517 @@
       "markers": [
         "pashto"
       ]
+    },
+    {
+      "id": "dataset-hf-aamirhs-pashto-audio-wav2vec",
+      "title": "aamirhs/pashto-audio-wav2vec",
+      "url": "https://huggingface.co/datasets/aamirhs/pashto-audio-wav2vec",
+      "category": "dataset",
+      "source": "huggingface",
+      "status": "verified",
+      "summary": "Pashto speech dataset surfaced from Hugging Face candidate sync for ASR experiments.",
+      "primary_use": "Pashto ASR data exploration and baseline training",
+      "tasks": [
+        "asr"
+      ],
+      "tags": [
+        "pashto",
+        "dataset",
+        "huggingface",
+        "speech",
+        "asr"
+      ],
+      "evidence_text": "Matched by Pashto keyword in Hugging Face search results.",
+      "evidence_url": "https://huggingface.co/datasets/aamirhs/pashto-audio-wav2vec",
+      "markers": [
+        "pashto"
+      ]
+    },
+    {
+      "id": "dataset-hf-alimuhammad73-pashto-poetry",
+      "title": "AliMuhammad73/Pashto-Poetry",
+      "url": "https://huggingface.co/datasets/AliMuhammad73/Pashto-Poetry",
+      "category": "dataset",
+      "source": "huggingface",
+      "status": "verified",
+      "summary": "Pashto poetry text dataset surfaced from Hugging Face candidate sync for NLP experiments.",
+      "primary_use": "Pashto poetry corpus for language modeling and text analysis",
+      "tasks": [
+        "nlp"
+      ],
+      "tags": [
+        "pashto",
+        "dataset",
+        "huggingface",
+        "text",
+        "poetry",
+        "nlp"
+      ],
+      "evidence_text": "Matched by Pashto keyword in Hugging Face search results.",
+      "evidence_url": "https://huggingface.co/datasets/AliMuhammad73/Pashto-Poetry",
+      "markers": [
+        "pashto"
+      ]
+    },
+    {
+      "id": "model-hf-aamirhs-wav2vec2-large-xls-r-300m-pashto-colab",
+      "title": "aamirhs/wav2vec2-large-xls-r-300m-pashto-colab",
+      "url": "https://huggingface.co/aamirhs/wav2vec2-large-xls-r-300m-pashto-colab",
+      "category": "model",
+      "source": "huggingface",
+      "status": "verified",
+      "summary": "Pashto ASR model checkpoint surfaced from Hugging Face candidate sync.",
+      "primary_use": "Pashto ASR baseline and transfer-learning comparison",
+      "tasks": [
+        "asr"
+      ],
+      "tags": [
+        "pashto",
+        "model",
+        "huggingface",
+        "asr"
+      ],
+      "evidence_text": "Matched by Pashto keyword in Hugging Face search results.",
+      "evidence_url": "https://huggingface.co/aamirhs/wav2vec2-large-xls-r-300m-pashto-colab",
+      "markers": [
+        "pashto"
+      ]
+    },
+    {
+      "id": "project-hf-space-aizazayyubi-pashto-asr",
+      "title": "Aizazayyubi/pashto_asr",
+      "url": "https://huggingface.co/spaces/Aizazayyubi/pashto_asr",
+      "category": "project",
+      "source": "huggingface",
+      "status": "verified",
+      "summary": "Pashto ASR interactive demo surfaced from Hugging Face Spaces candidate sync.",
+      "primary_use": "Interactive Pashto ASR demo for qualitative evaluation",
+      "tasks": [
+        "asr",
+        "demo"
+      ],
+      "tags": [
+        "pashto",
+        "project",
+        "huggingface",
+        "asr",
+        "demo"
+      ],
+      "evidence_text": "Matched by Pashto keyword in Hugging Face Spaces search.",
+      "evidence_url": "https://huggingface.co/spaces/Aizazayyubi/pashto_asr",
+      "markers": [
+        "pashto"
+      ]
+    },
+    {
+      "id": "paper-arxiv-from-scarcity-to-scale-pashto-common-voice",
+      "title": "From Scarcity to Scale: A Release-Level Analysis of the Pashto Common Voice Dataset",
+      "url": "http://arxiv.org/abs/2602.14062v1",
+      "category": "paper",
+      "source": "arxiv",
+      "status": "verified",
+      "summary": "Research paper analyzing Pashto Common Voice releases and dataset scaling characteristics.",
+      "primary_use": "ASR data quality and release trend reference",
+      "tasks": [
+        "asr",
+        "benchmarking"
+      ],
+      "tags": [
+        "pashto",
+        "paper",
+        "arxiv",
+        "asr",
+        "common-voice"
+      ],
+      "evidence_text": "Matched by Pashto marker in paper title from arXiv query results.",
+      "evidence_url": "http://arxiv.org/abs/2602.14062v1",
+      "markers": [
+        "pashto"
+      ]
+    },
+    {
+      "id": "paper-arxiv-tuning-traditional-pashto-text-classification",
+      "title": "Tuning Traditional Language Processing Approaches for Pashto Text Classification",
+      "url": "http://arxiv.org/abs/2305.03737v1",
+      "category": "paper",
+      "source": "arxiv",
+      "status": "verified",
+      "summary": "Research paper focused on Pashto text classification using traditional NLP approaches.",
+      "primary_use": "Pashto text classification method reference",
+      "tasks": [
+        "nlp"
+      ],
+      "tags": [
+        "pashto",
+        "paper",
+        "arxiv",
+        "nlp",
+        "classification"
+      ],
+      "evidence_text": "Matched by Pashto marker in paper title from arXiv query results.",
+      "evidence_url": "http://arxiv.org/abs/2305.03737v1",
+      "markers": [
+        "pashto"
+      ]
+    },
+    {
+      "id": "dataset-dataverse-iarpa-babel-pashto-language-pack-v0-4by",
+      "title": "IARPA Babel Pashto Language Pack IARPA-babel104b-v0.4bY",
+      "url": "https://hdl.handle.net/11272.1/AB2/GLFN3X",
+      "category": "dataset",
+      "source": "dataverse",
+      "status": "verified",
+      "summary": "Pashto Babel language pack dataset for speech and language processing evaluation.",
+      "primary_use": "Pashto speech dataset for ASR and language identification experiments",
+      "tasks": [
+        "asr",
+        "benchmarking"
+      ],
+      "tags": [
+        "pashto",
+        "dataset",
+        "dataverse",
+        "speech",
+        "asr",
+        "babel"
+      ],
+      "evidence_text": "Dataverse metadata includes Pashto markers in dataset title or description.",
+      "evidence_url": "https://hdl.handle.net/11272.1/AB2/GLFN3X",
+      "markers": [
+        "pashto"
+      ]
+    },
+    {
+      "id": "paper-arxiv-image-to-text-pashto-farsi-traditional-chinese",
+      "title": "Development of a New Image-to-text Conversion System for Pashto, Farsi and Traditional Chinese",
+      "url": "http://arxiv.org/abs/2005.08650v1",
+      "category": "paper",
+      "source": "arxiv",
+      "status": "verified",
+      "summary": "Research paper on image-to-text conversion including Pashto OCR.",
+      "primary_use": "Pashto OCR method reference",
+      "tasks": [
+        "ocr",
+        "nlp"
+      ],
+      "tags": [
+        "pashto",
+        "paper",
+        "arxiv",
+        "ocr"
+      ],
+      "evidence_text": "Matched by Pashto marker in paper title from arXiv query results.",
+      "evidence_url": "http://arxiv.org/abs/2005.08650v1",
+      "markers": [
+        "pashto"
+      ]
+    },
+    {
+      "id": "paper-openalex-benchmark-pashto-handwritten-character-dataset-ocr",
+      "title": "Benchmark Pashto Handwritten Character Dataset and Pashto Object Character Recognition (OCR) Using Deep Neural Network with Rule Activation Function",
+      "url": "https://doi.org/10.1155/2021/6669672",
+      "category": "paper",
+      "source": "openalex",
+      "status": "verified",
+      "summary": "Research paper introducing a benchmark dataset and OCR approach for Pashto handwritten characters.",
+      "primary_use": "Pashto handwritten OCR benchmark and methodology reference",
+      "tasks": [
+        "ocr",
+        "benchmarking"
+      ],
+      "tags": [
+        "pashto",
+        "paper",
+        "openalex",
+        "ocr",
+        "benchmark"
+      ],
+      "evidence_text": "Matched by explicit Pashto marker in title from OpenAlex works search.",
+      "evidence_url": "https://doi.org/10.1155/2021/6669672",
+      "markers": [
+        "pashto"
+      ]
+    },
+    {
+      "id": "paper-openalex-asr-isolated-pashto-spoken-digits-mfcc-knn",
+      "title": "Database development and automatic speech recognition of isolated Pashto spoken digits using MFCC and K-NN",
+      "url": "https://doi.org/10.1007/s10772-014-9267-z",
+      "category": "paper",
+      "source": "openalex",
+      "status": "verified",
+      "summary": "Research paper on isolated Pashto spoken-digit ASR with MFCC and K-NN.",
+      "primary_use": "Pashto ASR baseline method reference for digit recognition",
+      "tasks": [
+        "asr"
+      ],
+      "tags": [
+        "pashto",
+        "paper",
+        "openalex",
+        "asr",
+        "speech"
+      ],
+      "evidence_text": "Matched by explicit Pashto marker in title from OpenAlex works search.",
+      "evidence_url": "https://doi.org/10.1007/s10772-014-9267-z",
+      "markers": [
+        "pashto"
+      ]
+    },
+    {
+      "id": "paper-openalex-pashto-isolated-digits-recognition-dcnn",
+      "title": "Pashto isolated digits recognition using deep convolutional neural network",
+      "url": "https://doi.org/10.1016/j.heliyon.2020.e03372",
+      "category": "paper",
+      "source": "openalex",
+      "status": "verified",
+      "summary": "Research paper on Pashto isolated-digit recognition using deep convolutional neural networks.",
+      "primary_use": "Pashto speech recognition research reference",
+      "tasks": [
+        "asr"
+      ],
+      "tags": [
+        "pashto",
+        "paper",
+        "openalex",
+        "asr",
+        "deep-learning"
+      ],
+      "evidence_text": "Matched by explicit Pashto marker in title from OpenAlex works search.",
+      "evidence_url": "https://doi.org/10.1016/j.heliyon.2020.e03372",
+      "markers": [
+        "pashto"
+      ]
+    },
+    {
+      "id": "paper-openalex-pashto-offensive-language-detection-benchmark-bert",
+      "title": "Pashto offensive language detection: a benchmark dataset and monolingual Pashto BERT",
+      "url": "https://doi.org/10.7717/peerj-cs.1617",
+      "category": "paper",
+      "source": "openalex",
+      "status": "verified",
+      "summary": "Research paper on Pashto offensive language detection with benchmark dataset and monolingual BERT model.",
+      "primary_use": "Pashto NLP toxicity detection benchmark and model reference",
+      "tasks": [
+        "nlp",
+        "benchmarking"
+      ],
+      "tags": [
+        "pashto",
+        "paper",
+        "openalex",
+        "nlp",
+        "bert",
+        "benchmark"
+      ],
+      "evidence_text": "Matched by explicit Pashto marker in title from OpenAlex works search.",
+      "evidence_url": "https://doi.org/10.7717/peerj-cs.1617",
+      "markers": [
+        "pashto"
+      ]
+    },
+    {
+      "id": "paper-openalex-phti-pashto-handwritten-text-imagebase",
+      "title": "PHTI: Pashto Handwritten Text Imagebase for Deep Learning Applications",
+      "url": "https://doi.org/10.1109/access.2022.3216881",
+      "category": "paper",
+      "source": "openalex",
+      "status": "verified",
+      "summary": "Research paper describing PHTI, a Pashto handwritten text imagebase for deep learning.",
+      "primary_use": "Pashto OCR dataset and benchmark reference",
+      "tasks": [
+        "ocr",
+        "benchmarking"
+      ],
+      "tags": [
+        "pashto",
+        "paper",
+        "openalex",
+        "ocr",
+        "dataset"
+      ],
+      "evidence_text": "Matched by explicit Pashto marker in title from OpenAlex works search.",
+      "evidence_url": "https://doi.org/10.1109/access.2022.3216881",
+      "markers": [
+        "pashto"
+      ]
+    },
+    {
+      "id": "paper-openalex-recognition-of-pashto-handwritten-characters-deep-learning",
+      "title": "Recognition of Pashto Handwritten Characters Based on Deep Learning",
+      "url": "https://doi.org/10.3390/s20205884",
+      "category": "paper",
+      "source": "openalex",
+      "status": "verified",
+      "summary": "Research paper on deep-learning-based recognition of Pashto handwritten characters.",
+      "primary_use": "Pashto OCR model reference for handwritten character recognition",
+      "tasks": [
+        "ocr"
+      ],
+      "tags": [
+        "pashto",
+        "paper",
+        "openalex",
+        "ocr",
+        "deep-learning"
+      ],
+      "evidence_text": "Matched by explicit Pashto marker in title from OpenAlex works search.",
+      "evidence_url": "https://doi.org/10.3390/s20205884",
+      "markers": [
+        "pashto"
+      ]
+    },
+    {
+      "id": "paper-openalex-kpti-katib-pashto-text-imagebase-benchmark",
+      "title": "KPTI: Katib's Pashto Text Imagebase and Deep Learning Benchmark",
+      "url": "https://doi.org/10.1109/icfhr.2016.0090",
+      "category": "paper",
+      "source": "openalex",
+      "status": "verified",
+      "summary": "Research paper introducing KPTI, a Pashto text imagebase and benchmark for handwritten recognition.",
+      "primary_use": "Pashto OCR dataset and benchmarking reference",
+      "tasks": [
+        "ocr",
+        "benchmarking"
+      ],
+      "tags": [
+        "pashto",
+        "paper",
+        "openalex",
+        "ocr",
+        "benchmark"
+      ],
+      "evidence_text": "Matched by explicit Pashto marker in title from OpenAlex works search.",
+      "evidence_url": "https://doi.org/10.1109/icfhr.2016.0090",
+      "markers": [
+        "pashto"
+      ]
+    },
+    {
+      "id": "paper-openalex-pioneer-dataset-handwritten-pashto-cnn",
+      "title": "Pioneer dataset and recognition of Handwritten Pashto characters using Convolution Neural Networks",
+      "url": "https://doi.org/10.1177/0020294020964826",
+      "category": "paper",
+      "source": "openalex",
+      "status": "verified",
+      "summary": "Research paper on a pioneer handwritten Pashto character dataset with CNN-based recognition.",
+      "primary_use": "Pashto handwritten character recognition reference",
+      "tasks": [
+        "ocr",
+        "benchmarking"
+      ],
+      "tags": [
+        "pashto",
+        "paper",
+        "openalex",
+        "ocr",
+        "deep-learning"
+      ],
+      "evidence_text": "Matched by explicit Pashto marker in title from OpenAlex works search.",
+      "evidence_url": "https://doi.org/10.1177/0020294020964826",
+      "markers": [
+        "pashto"
+      ]
+    },
+    {
+      "id": "paper-openalex-scale-rotation-invariant-ocr-pashto-mdlstm",
+      "title": "Scale and rotation invariant OCR for Pashto cursive script using MDLSTM network",
+      "url": "https://doi.org/10.1109/icdar.2015.7333931",
+      "category": "paper",
+      "source": "openalex",
+      "status": "verified",
+      "summary": "Research paper on scale- and rotation-invariant OCR for cursive Pashto using MDLSTM.",
+      "primary_use": "Pashto OCR model architecture reference",
+      "tasks": [
+        "ocr"
+      ],
+      "tags": [
+        "pashto",
+        "paper",
+        "openalex",
+        "ocr",
+        "mdlstm"
+      ],
+      "evidence_text": "Matched by explicit Pashto marker in title from OpenAlex works search.",
+      "evidence_url": "https://doi.org/10.1109/icdar.2015.7333931",
+      "markers": [
+        "pashto"
+      ]
+    },
+    {
+      "id": "paper-openalex-recognizable-units-pashto-ocr",
+      "title": "Recognizable units in Pashto language for OCR",
+      "url": "https://doi.org/10.1109/icdar.2015.7333963",
+      "category": "paper",
+      "source": "openalex",
+      "status": "verified",
+      "summary": "Research paper defining recognizable units in Pashto for OCR workflows.",
+      "primary_use": "Pashto OCR preprocessing and unit-design reference",
+      "tasks": [
+        "ocr"
+      ],
+      "tags": [
+        "pashto",
+        "paper",
+        "openalex",
+        "ocr"
+      ],
+      "evidence_text": "Matched by explicit Pashto marker in title from OpenAlex works search.",
+      "evidence_url": "https://doi.org/10.1109/icdar.2015.7333963",
+      "markers": [
+        "pashto"
+      ]
+    },
+    {
+      "id": "paper-openalex-shape-analysis-pashto-script-image-database-ocr",
+      "title": "Shape analysis of Pashto script and creation of image database for OCR",
+      "url": "https://doi.org/10.1109/icet.2009.5353160",
+      "category": "paper",
+      "source": "openalex",
+      "status": "verified",
+      "summary": "Research paper on Pashto script shape analysis and image database creation for OCR.",
+      "primary_use": "Pashto OCR dataset design and feature reference",
+      "tasks": [
+        "ocr",
+        "benchmarking"
+      ],
+      "tags": [
+        "pashto",
+        "paper",
+        "openalex",
+        "ocr",
+        "dataset"
+      ],
+      "evidence_text": "Matched by explicit Pashto marker in title from OpenAlex works search.",
+      "evidence_url": "https://doi.org/10.1109/icet.2009.5353160",
+      "markers": [
+        "pashto"
+      ]
+    },
+    {
+      "id": "paper-openalex-speech-translation-low-resource-case-pashto",
+      "title": "Speech translation for low-resource languages: the case of Pashto",
+      "url": "https://doi.org/10.21437/interspeech.2005-723",
+      "category": "paper",
+      "source": "openalex",
+      "status": "verified",
+      "summary": "Research paper on speech translation for low-resource languages, including Pashto.",
+      "primary_use": "Pashto speech translation and low-resource MT reference",
+      "tasks": [
+        "asr",
+        "mt"
+      ],
+      "tags": [
+        "pashto",
+        "paper",
+        "openalex",
+        "speech",
+        "translation"
+      ],
+      "evidence_text": "Matched by explicit Pashto marker in title from OpenAlex works search.",
+      "evidence_url": "https://doi.org/10.21437/interspeech.2005-723",
+      "markers": [
+        "pashto"
+      ]
     }
   ]
 }

pyproject.toml

@@ -7,6 +7,22 @@ name = "pukhto-pashto"
 version = "0.1.0"
 description = "Open Pashto language resources for ASR, TTS, NLP, and benchmarks"
 requires-python = ">=3.10"
+readme = "README.md"
+keywords = [
+  "pashto",
+  "pukhto",
+  "pushto",
+  "asr",
+  "tts",
+  "nlp",
+  "machine-translation",
+  "language-resources",
+]
+
+[project.urls]
+Homepage = "https://github.com/Musawer1214/pashto-language-resources"
+Documentation = "https://musawer1214.github.io/pashto-language-resources/"
+Repository = "https://github.com/Musawer1214/pashto-language-resources"
 
 [project.optional-dependencies]
 dev = ["pytest>=8.0.0"]
@@ -17,3 +33,4 @@ python_files = ["test_*.py"]
 
 [tool.setuptools]
 packages = []
+

resources/README.md

@@ -3,12 +3,12 @@
 Structured, Pashto-focused resource tracking lives in this folder.
 
 ## Sections
-- Datasets (32): [datasets/README.md](datasets/README.md)
-- Models (15): [models/README.md](models/README.md)
+- Datasets (35): [datasets/README.md](datasets/README.md)
+- Models (16): [models/README.md](models/README.md)
 - Benchmarks (4): [benchmarks/README.md](benchmarks/README.md)
 - Tools (0): [tools/README.md](tools/README.md)
-- Papers (9): [papers/README.md](papers/README.md)
-- Projects (14): [projects/README.md](projects/README.md)
+- Papers (24): [papers/README.md](papers/README.md)
+- Projects (15): [projects/README.md](projects/README.md)
 - Code (1): [codes/README.md](codes/README.md)
 
 ## Machine-Readable Catalog
@@ -22,4 +22,4 @@ Structured, Pashto-focused resource tracking lives in this folder.
 - Run `python scripts/validate_resource_catalog.py` before opening a PR.
 - Run `python scripts/generate_resource_views.py` after catalog changes.
 
-Verified resource count: `75`
+Verified resource count: `95`

resources/catalog/resources.json

@@ -1,6 +1,6 @@
 {
   "version": "1.0.0",
-  "updated_on": "2026-02-16",
+  "updated_on": "2026-02-17",
   "resources": [
     {
       "id": "dataset-common-voice-ps-v24",
@@ -2008,6 +2008,557 @@
         "other",
         "nlp"
       ]
+    },
+    {
+      "id": "dataset-hf-aamirhs-pashto-audio-wav2vec",
+      "title": "aamirhs/pashto-audio-wav2vec",
+      "url": "https://huggingface.co/datasets/aamirhs/pashto-audio-wav2vec",
+      "category": "dataset",
+      "source": "huggingface",
+      "status": "verified",
+      "summary": "Pashto speech dataset surfaced from Hugging Face candidate sync for ASR experiments.",
+      "primary_use": "Pashto ASR data exploration and baseline training",
+      "tasks": [
+        "asr"
+      ],
+      "pashto_evidence": {
+        "evidence_text": "Matched by Pashto keyword in Hugging Face search results.",
+        "evidence_url": "https://huggingface.co/datasets/aamirhs/pashto-audio-wav2vec",
+        "markers": [
+          "pashto"
+        ]
+      },
+      "tags": [
+        "pashto",
+        "dataset",
+        "huggingface",
+        "speech",
+        "asr"
+      ]
+    },
+    {
+      "id": "dataset-hf-alimuhammad73-pashto-poetry",
+      "title": "AliMuhammad73/Pashto-Poetry",
+      "url": "https://huggingface.co/datasets/AliMuhammad73/Pashto-Poetry",
+      "category": "dataset",
+      "source": "huggingface",
+      "status": "verified",
+      "summary": "Pashto poetry text dataset surfaced from Hugging Face candidate sync for NLP experiments.",
+      "primary_use": "Pashto poetry corpus for language modeling and text analysis",
+      "tasks": [
+        "nlp"
+      ],
+      "pashto_evidence": {
+        "evidence_text": "Matched by Pashto keyword in Hugging Face search results.",
+        "evidence_url": "https://huggingface.co/datasets/AliMuhammad73/Pashto-Poetry",
+        "markers": [
+          "pashto"
+        ]
+      },
+      "tags": [
+        "pashto",
+        "dataset",
+        "huggingface",
+        "text",
+        "poetry",
+        "nlp"
+      ]
+    },
+    {
+      "id": "model-hf-aamirhs-wav2vec2-large-xls-r-300m-pashto-colab",
+      "title": "aamirhs/wav2vec2-large-xls-r-300m-pashto-colab",
+      "url": "https://huggingface.co/aamirhs/wav2vec2-large-xls-r-300m-pashto-colab",
+      "category": "model",
+      "source": "huggingface",
+      "status": "verified",
+      "summary": "Pashto ASR model checkpoint surfaced from Hugging Face candidate sync.",
+      "primary_use": "Pashto ASR baseline and transfer-learning comparison",
+      "tasks": [
+        "asr"
+      ],
+      "pashto_evidence": {
+        "evidence_text": "Matched by Pashto keyword in Hugging Face search results.",
+        "evidence_url": "https://huggingface.co/aamirhs/wav2vec2-large-xls-r-300m-pashto-colab",
+        "markers": [
+          "pashto"
+        ]
+      },
+      "tags": [
+        "pashto",
+        "model",
+        "huggingface",
+        "asr"
+      ]
+    },
+    {
+      "id": "project-hf-space-aizazayyubi-pashto-asr",
+      "title": "Aizazayyubi/pashto_asr",
+      "url": "https://huggingface.co/spaces/Aizazayyubi/pashto_asr",
+      "category": "project",
+      "source": "huggingface",
+      "status": "verified",
+      "summary": "Pashto ASR interactive demo surfaced from Hugging Face Spaces candidate sync.",
+      "primary_use": "Interactive Pashto ASR demo for qualitative evaluation",
+      "tasks": [
+        "asr",
+        "demo"
+      ],
+      "pashto_evidence": {
+        "evidence_text": "Matched by Pashto keyword in Hugging Face Spaces search.",
+        "evidence_url": "https://huggingface.co/spaces/Aizazayyubi/pashto_asr",
+        "markers": [
+          "pashto"
+        ]
+      },
+      "tags": [
+        "pashto",
+        "project",
+        "huggingface",
+        "asr",
+        "demo"
+      ]
+    },
+    {
+      "id": "paper-arxiv-from-scarcity-to-scale-pashto-common-voice",
+      "title": "From Scarcity to Scale: A Release-Level Analysis of the Pashto Common Voice Dataset",
+      "url": "http://arxiv.org/abs/2602.14062v1",
+      "category": "paper",
+      "source": "arxiv",
+      "status": "verified",
+      "summary": "Research paper analyzing Pashto Common Voice releases and dataset scaling characteristics.",
+      "primary_use": "ASR data quality and release trend reference",
+      "tasks": [
+        "asr",
+        "benchmarking"
+      ],
+      "pashto_evidence": {
+        "evidence_text": "Matched by Pashto marker in paper title from arXiv query results.",
+        "evidence_url": "http://arxiv.org/abs/2602.14062v1",
+        "markers": [
+          "pashto"
+        ]
+      },
+      "tags": [
+        "pashto",
+        "paper",
+        "arxiv",
+        "asr",
+        "common-voice"
+      ]
+    },
+    {
+      "id": "paper-arxiv-tuning-traditional-pashto-text-classification",
+      "title": "Tuning Traditional Language Processing Approaches for Pashto Text Classification",
+      "url": "http://arxiv.org/abs/2305.03737v1",
+      "category": "paper",
+      "source": "arxiv",
+      "status": "verified",
+      "summary": "Research paper focused on Pashto text classification using traditional NLP approaches.",
+      "primary_use": "Pashto text classification method reference",
+      "tasks": [
+        "nlp"
+      ],
+      "pashto_evidence": {
+        "evidence_text": "Matched by Pashto marker in paper title from arXiv query results.",
+        "evidence_url": "http://arxiv.org/abs/2305.03737v1",
+        "markers": [
+          "pashto"
+        ]
+      },
+      "tags": [
+        "pashto",
+        "paper",
+        "arxiv",
+        "nlp",
+        "classification"
+      ]
+    },
+    {
+      "id": "dataset-dataverse-iarpa-babel-pashto-language-pack-v0-4by",
+      "title": "IARPA Babel Pashto Language Pack IARPA-babel104b-v0.4bY",
+      "url": "https://hdl.handle.net/11272.1/AB2/GLFN3X",
+      "category": "dataset",
+      "source": "dataverse",
+      "status": "verified",
+      "summary": "Pashto Babel language pack dataset for speech and language processing evaluation.",
+      "primary_use": "Pashto speech dataset for ASR and language identification experiments",
+      "tasks": [
+        "asr",
+        "benchmarking"
+      ],
+      "pashto_evidence": {
+        "evidence_text": "Dataverse metadata includes Pashto markers in dataset title or description.",
+        "evidence_url": "https://hdl.handle.net/11272.1/AB2/GLFN3X",
+        "markers": [
+          "pashto"
+        ]
+      },
+      "tags": [
+        "pashto",
+        "dataset",
+        "dataverse",
+        "speech",
+        "asr",
+        "babel"
+      ]
+    },
+    {
+      "id": "paper-arxiv-image-to-text-pashto-farsi-traditional-chinese",
+      "title": "Development of a New Image-to-text Conversion System for Pashto, Farsi and Traditional Chinese",
+      "url": "http://arxiv.org/abs/2005.08650v1",
+      "category": "paper",
+      "source": "arxiv",
+      "status": "verified",
+      "summary": "Research paper on image-to-text conversion including Pashto OCR.",
+      "primary_use": "Pashto OCR method reference",
+      "tasks": [
+        "ocr",
+        "nlp"
+      ],
+      "pashto_evidence": {
+        "evidence_text": "Matched by Pashto marker in paper title from arXiv query results.",
+        "evidence_url": "http://arxiv.org/abs/2005.08650v1",
+        "markers": [
+          "pashto"
+        ]
+      },
+      "tags": [
+        "pashto",
+        "paper",
+        "arxiv",
+        "ocr"
+      ]
+    },
+    {
+      "id": "paper-openalex-benchmark-pashto-handwritten-character-dataset-ocr",
+      "title": "Benchmark Pashto Handwritten Character Dataset and Pashto Object Character Recognition (OCR) Using Deep Neural Network with Rule Activation Function",
+      "url": "https://doi.org/10.1155/2021/6669672",
+      "category": "paper",
+      "source": "openalex",
+      "status": "verified",
+      "summary": "Research paper introducing a benchmark dataset and OCR approach for Pashto handwritten characters.",
+      "primary_use": "Pashto handwritten OCR benchmark and methodology reference",
+      "tasks": [
+        "ocr",
+        "benchmarking"
+      ],
+      "pashto_evidence": {
+        "evidence_text": "Matched by explicit Pashto marker in title from OpenAlex works search.",
+        "evidence_url": "https://doi.org/10.1155/2021/6669672",
+        "markers": [
+          "pashto"
+        ]
+      },
+      "tags": [
+        "pashto",
+        "paper",
+        "openalex",
+        "ocr",
+        "benchmark"
+      ]
+    },
+    {
+      "id": "paper-openalex-asr-isolated-pashto-spoken-digits-mfcc-knn",
+      "title": "Database development and automatic speech recognition of isolated Pashto spoken digits using MFCC and K-NN",
+      "url": "https://doi.org/10.1007/s10772-014-9267-z",
+      "category": "paper",
+      "source": "openalex",
+      "status": "verified",
+      "summary": "Research paper on isolated Pashto spoken-digit ASR with MFCC and K-NN.",
+      "primary_use": "Pashto ASR baseline method reference for digit recognition",
+      "tasks": [
+        "asr"
+      ],
+      "pashto_evidence": {
+        "evidence_text": "Matched by explicit Pashto marker in title from OpenAlex works search.",
+        "evidence_url": "https://doi.org/10.1007/s10772-014-9267-z",
+        "markers": [
+          "pashto"
+        ]
+      },
+      "tags": [
+        "pashto",
+        "paper",
+        "openalex",
+        "asr",
+        "speech"
+      ]
+    },
+    {
+      "id": "paper-openalex-pashto-isolated-digits-recognition-dcnn",
+      "title": "Pashto isolated digits recognition using deep convolutional neural network",
+      "url": "https://doi.org/10.1016/j.heliyon.2020.e03372",
+      "category": "paper",
+      "source": "openalex",
+      "status": "verified",
+      "summary": "Research paper on Pashto isolated-digit recognition using deep convolutional neural networks.",
+      "primary_use": "Pashto speech recognition research reference",
+      "tasks": [
+        "asr"
+      ],
+      "pashto_evidence": {
+        "evidence_text": "Matched by explicit Pashto marker in title from OpenAlex works search.",
+        "evidence_url": "https://doi.org/10.1016/j.heliyon.2020.e03372",
+        "markers": [
+          "pashto"
+        ]
+      },
+      "tags": [
+        "pashto",
+        "paper",
+        "openalex",
+        "asr",
+        "deep-learning"
+      ]
+    },
+    {
+      "id": "paper-openalex-pashto-offensive-language-detection-benchmark-bert",
+      "title": "Pashto offensive language detection: a benchmark dataset and monolingual Pashto BERT",
+      "url": "https://doi.org/10.7717/peerj-cs.1617",
+      "category": "paper",
+      "source": "openalex",
+      "status": "verified",
+      "summary": "Research paper on Pashto offensive language detection with benchmark dataset and monolingual BERT model.",
+      "primary_use": "Pashto NLP toxicity detection benchmark and model reference",
+      "tasks": [
+        "nlp",
+        "benchmarking"
+      ],
+      "pashto_evidence": {
+        "evidence_text": "Matched by explicit Pashto marker in title from OpenAlex works search.",
+        "evidence_url": "https://doi.org/10.7717/peerj-cs.1617",
+        "markers": [
+          "pashto"
+        ]
+      },
+      "tags": [
+        "pashto",
+        "paper",
+        "openalex",
+        "nlp",
+        "bert",
+        "benchmark"
+      ]
+    },
+    {
+      "id": "paper-openalex-phti-pashto-handwritten-text-imagebase",
+      "title": "PHTI: Pashto Handwritten Text Imagebase for Deep Learning Applications",
+      "url": "https://doi.org/10.1109/access.2022.3216881",
+      "category": "paper",
+      "source": "openalex",
+      "status": "verified",
+      "summary": "Research paper describing PHTI, a Pashto handwritten text imagebase for deep learning.",
+      "primary_use": "Pashto OCR dataset and benchmark reference",
+      "tasks": [
+        "ocr",
+        "benchmarking"
+      ],
+      "pashto_evidence": {
+        "evidence_text": "Matched by explicit Pashto marker in title from OpenAlex works search.",
+        "evidence_url": "https://doi.org/10.1109/access.2022.3216881",
+        "markers": [
+          "pashto"
+        ]
+      },
+      "tags": [
+        "pashto",
+        "paper",
+        "openalex",
+        "ocr",
+        "dataset"
+      ]
+    },
+    {
+      "id": "paper-openalex-recognition-of-pashto-handwritten-characters-deep-learning",
+      "title": "Recognition of Pashto Handwritten Characters Based on Deep Learning",
+      "url": "https://doi.org/10.3390/s20205884",
+      "category": "paper",
+      "source": "openalex",
+      "status": "verified",
+      "summary": "Research paper on deep-learning-based recognition of Pashto handwritten characters.",
+      "primary_use": "Pashto OCR model reference for handwritten character recognition",
+      "tasks": [
+        "ocr"
+      ],
+      "pashto_evidence": {
+        "evidence_text": "Matched by explicit Pashto marker in title from OpenAlex works search.",
+        "evidence_url": "https://doi.org/10.3390/s20205884",
+        "markers": [
+          "pashto"
+        ]
+      },
+      "tags": [
+        "pashto",
+        "paper",
+        "openalex",
+        "ocr",
+        "deep-learning"
+      ]
+    },
+    {
+      "id": "paper-openalex-kpti-katib-pashto-text-imagebase-benchmark",
+      "title": "KPTI: Katib's Pashto Text Imagebase and Deep Learning Benchmark",
+      "url": "https://doi.org/10.1109/icfhr.2016.0090",
+      "category": "paper",
+      "source": "openalex",
+      "status": "verified",
+      "summary": "Research paper introducing KPTI, a Pashto text imagebase and benchmark for handwritten recognition.",
+      "primary_use": "Pashto OCR dataset and benchmarking reference",
+      "tasks": [
+        "ocr",
+        "benchmarking"
+      ],
+      "pashto_evidence": {
+        "evidence_text": "Matched by explicit Pashto marker in title from OpenAlex works search.",
+        "evidence_url": "https://doi.org/10.1109/icfhr.2016.0090",
+        "markers": [
+          "pashto"
+        ]
+      },
+      "tags": [
+        "pashto",
+        "paper",
+        "openalex",
+        "ocr",
+        "benchmark"
+      ]
+    },
+    {
+      "id": "paper-openalex-pioneer-dataset-handwritten-pashto-cnn",
+      "title": "Pioneer dataset and recognition of Handwritten Pashto characters using Convolution Neural Networks",
+      "url": "https://doi.org/10.1177/0020294020964826",
+      "category": "paper",
+      "source": "openalex",
+      "status": "verified",
+      "summary": "Research paper on a pioneer handwritten Pashto character dataset with CNN-based recognition.",
+      "primary_use": "Pashto handwritten character recognition reference",
+      "tasks": [
+        "ocr",
+        "benchmarking"
+      ],
+      "pashto_evidence": {
+        "evidence_text": "Matched by explicit Pashto marker in title from OpenAlex works search.",
+        "evidence_url": "https://doi.org/10.1177/0020294020964826",
+        "markers": [
+          "pashto"
+        ]
+      },
+      "tags": [
+        "pashto",
+        "paper",
+        "openalex",
+        "ocr",
+        "deep-learning"
+      ]
+    },
+    {
+      "id": "paper-openalex-scale-rotation-invariant-ocr-pashto-mdlstm",
+      "title": "Scale and rotation invariant OCR for Pashto cursive script using MDLSTM network",
+      "url": "https://doi.org/10.1109/icdar.2015.7333931",
+      "category": "paper",
+      "source": "openalex",
+      "status": "verified",
+      "summary": "Research paper on scale- and rotation-invariant OCR for cursive Pashto using MDLSTM.",
+      "primary_use": "Pashto OCR model architecture reference",
+      "tasks": [
+        "ocr"
+      ],
+      "pashto_evidence": {
+        "evidence_text": "Matched by explicit Pashto marker in title from OpenAlex works search.",
+        "evidence_url": "https://doi.org/10.1109/icdar.2015.7333931",
+        "markers": [
+          "pashto"
+        ]
+      },
+      "tags": [
+        "pashto",
+        "paper",
+        "openalex",
+        "ocr",
+        "mdlstm"
+      ]
+    },
+    {
+      "id": "paper-openalex-recognizable-units-pashto-ocr",
+      "title": "Recognizable units in Pashto language for OCR",
+      "url": "https://doi.org/10.1109/icdar.2015.7333963",
+      "category": "paper",
+      "source": "openalex",
+      "status": "verified",
+      "summary": "Research paper defining recognizable units in Pashto for OCR workflows.",
+      "primary_use": "Pashto OCR preprocessing and unit-design reference",
+      "tasks": [
+        "ocr"
+      ],
+      "pashto_evidence": {
+        "evidence_text": "Matched by explicit Pashto marker in title from OpenAlex works search.",
+        "evidence_url": "https://doi.org/10.1109/icdar.2015.7333963",
+        "markers": [
+          "pashto"
+        ]
+      },
+      "tags": [
+        "pashto",
+        "paper",
+        "openalex",
+        "ocr"
+      ]
+    },
+    {
+      "id": "paper-openalex-shape-analysis-pashto-script-image-database-ocr",
+      "title": "Shape analysis of Pashto script and creation of image database for OCR",
+      "url": "https://doi.org/10.1109/icet.2009.5353160",
+      "category": "paper",
+      "source": "openalex",
+      "status": "verified",
+      "summary": "Research paper on Pashto script shape analysis and image database creation for OCR.",
+      "primary_use": "Pashto OCR dataset design and feature reference",
+      "tasks": [
+        "ocr",
+        "benchmarking"
+      ],
+      "pashto_evidence": {
+        "evidence_text": "Matched by explicit Pashto marker in title from OpenAlex works search.",
+        "evidence_url": "https://doi.org/10.1109/icet.2009.5353160",
+        "markers": [
+          "pashto"
+        ]
+      },
+      "tags": [
+        "pashto",
+        "paper",
+        "openalex",
+        "ocr",
+        "dataset"
+      ]
+    },
+    {
+      "id": "paper-openalex-speech-translation-low-resource-case-pashto",
+      "title": "Speech translation for low-resource languages: the case of Pashto",
+      "url": "https://doi.org/10.21437/interspeech.2005-723",
+      "category": "paper",
+      "source": "openalex",
+      "status": "verified",
+      "summary": "Research paper on speech translation for low-resource languages, including Pashto.",
+      "primary_use": "Pashto speech translation and low-resource MT reference",
+      "tasks": [
+        "asr",
+        "mt"
+      ],
+      "pashto_evidence": {
+        "evidence_text": "Matched by explicit Pashto marker in title from OpenAlex works search.",
+        "evidence_url": "https://doi.org/10.21437/interspeech.2005-723",
+        "markers": [
+          "pashto"
+        ]
+      },
+      "tags": [
+        "pashto",
+        "paper",
+        "openalex",
+        "speech",
+        "translation"
+      ]
     }
   ]
 }

resources/datasets/README.md

@@ -5,7 +5,9 @@
 | Resource | Link | Pashto Evidence | Primary Use |
 |---|---|---|---|
 | 99 Hours Pashto Spontaneous Dialogue Smartphone Speech Dataset | [huggingface](https://huggingface.co/datasets/Nexdata/99_Hours_Pashto_Spontaneous_Dialogue_Smartphone_speech_dataset) | [Dataset title explicitly includes Pashto and API metadata marks audio and text modalities. (`Pashto`)](https://huggingface.co/datasets/Nexdata/99_Hours_Pashto_Spontaneous_Dialogue_Smartphone_speech_dataset) | Spontaneous speech ASR training and robustness evaluation |
+| aamirhs/pashto-audio-wav2vec | [huggingface](https://huggingface.co/datasets/aamirhs/pashto-audio-wav2vec) | [Matched by Pashto keyword in Hugging Face search results. (`pashto`)](https://huggingface.co/datasets/aamirhs/pashto-audio-wav2vec) | Pashto ASR data exploration and baseline training |
 | adnankhan769/proper_dataset_english_2_pashto | [huggingface](https://huggingface.co/datasets/adnankhan769/proper_dataset_english_2_pashto) | [Matched by Pashto keyword in Hugging Face search results. (`pashto`)](https://huggingface.co/datasets/adnankhan769/proper_dataset_english_2_pashto) | Machine translation and bilingual corpus development |
+| AliMuhammad73/Pashto-Poetry | [huggingface](https://huggingface.co/datasets/AliMuhammad73/Pashto-Poetry) | [Matched by Pashto keyword in Hugging Face search results. (`pashto`)](https://huggingface.co/datasets/AliMuhammad73/Pashto-Poetry) | Pashto poetry corpus for language modeling and text analysis |
 | alpaca-pashto-cleaned | [huggingface](https://huggingface.co/datasets/saillab/alpaca-pashto-cleaned) | [Dataset metadata includes language:ps and dataset name includes Pashto. (`ps`, `Pashto`)](https://huggingface.co/api/datasets/saillab/alpaca-pashto-cleaned) | Pashto instruction tuning and conversational NLP experiments |
 | Belebele | [huggingface](https://huggingface.co/datasets/facebook/belebele) | [Dataset includes pbt_Arab subset. (`pbt_Arab`)](https://huggingface.co/datasets/facebook/belebele) | Comprehension and multilingual NLP benchmark |
 | Common Voice 24.0: Pashto Speech Dataset | [kaggle](https://www.kaggle.com/datasets/ataullahaali/common-voice-scripted-speech-24-0-pashto) | [Kaggle dataset title/subtitle includes Pashto keyword. (`Pashto`)](https://www.kaggle.com/datasets/ataullahaali/common-voice-scripted-speech-24-0-pashto) | ASR training and evaluation data source |
@@ -13,6 +15,7 @@
 | English to Pashto Sentences Dataset | [huggingface](https://huggingface.co/datasets/adnankhan769/english_to_pashto_sentences_dataset) | [Dataset ID explicitly states English-to-Pashto and includes Pashto-script sentence column. (`Pashto`)](https://huggingface.co/api/datasets/adnankhan769/english_to_pashto_sentences_dataset) | MT and bilingual sentence alignment baseline |
 | English-Pashto Language Dataset (EPLD) | [kaggle](https://www.kaggle.com/datasets/rabiakhan827/english-pashto-language-dataset-epld) | [Kaggle dataset title/subtitle includes Pashto keyword. (`Pashto`)](https://www.kaggle.com/datasets/rabiakhan827/english-pashto-language-dataset-epld) | Machine translation and bilingual corpus development |
 | Google FLEURS | [huggingface](https://huggingface.co/datasets/google/fleurs) | [Dataset config includes ps_af. (`ps_af`)](https://huggingface.co/datasets/google/fleurs/blob/main/fleurs.py) | Speech benchmark and external evaluation |
+| IARPA Babel Pashto Language Pack IARPA-babel104b-v0.4bY | [dataverse](https://hdl.handle.net/11272.1/AB2/GLFN3X) | [Dataverse metadata includes Pashto markers in dataset title or description. (`pashto`)](https://hdl.handle.net/11272.1/AB2/GLFN3X) | Pashto speech dataset for ASR and language identification experiments |
 | ihanif/pashto_asr_wer | [huggingface](https://huggingface.co/datasets/ihanif/pashto_asr_wer) | [Matched by Pashto keyword in Hugging Face search results. (`pashto`)](https://huggingface.co/datasets/ihanif/pashto_asr_wer) | ASR training and evaluation data source |
 | ihanif/pashto_speech_20k | [huggingface](https://huggingface.co/datasets/ihanif/pashto_speech_20k) | [Matched by Pashto keyword in Hugging Face search results. (`pashto`)](https://huggingface.co/datasets/ihanif/pashto_speech_20k) | ASR training and evaluation data source |
 | ihanif/pashto_speech_5k | [huggingface](https://huggingface.co/datasets/ihanif/pashto_speech_5k) | [Matched by Pashto keyword in Hugging Face search results. (`pashto`)](https://huggingface.co/datasets/ihanif/pashto_speech_5k) | ASR training and evaluation data source |

resources/models/README.md

@@ -4,6 +4,7 @@
 
 | Resource | Link | Pashto Evidence | Primary Use |
 |---|---|---|---|
+| aamirhs/wav2vec2-large-xls-r-300m-pashto-colab | [huggingface](https://huggingface.co/aamirhs/wav2vec2-large-xls-r-300m-pashto-colab) | [Matched by Pashto keyword in Hugging Face search results. (`pashto`)](https://huggingface.co/aamirhs/wav2vec2-large-xls-r-300m-pashto-colab) | Pashto ASR baseline and transfer-learning comparison |
 | ihanif/pashto-asr-base | [huggingface](https://huggingface.co/ihanif/pashto-asr-base) | [Matched by Pashto keyword in Hugging Face search results. (`pashto`)](https://huggingface.co/ihanif/pashto-asr-base) | Pashto ASR baseline and model comparison |
 | ihanif/wav2vec2-xls-r-300m-pashto-lm | [huggingface](https://huggingface.co/ihanif/wav2vec2-xls-r-300m-pashto-lm) | [Matched by Pashto keyword in Hugging Face search results. (`pashto`)](https://huggingface.co/ihanif/wav2vec2-xls-r-300m-pashto-lm) | Pashto ASR baseline and model comparison |
 | ihanif/whisper-large-pashto | [huggingface](https://huggingface.co/ihanif/whisper-large-pashto) | [Matched by Pashto keyword in Hugging Face search results. (`pashto`)](https://huggingface.co/ihanif/whisper-large-pashto) | Pashto ASR baseline and model comparison |

resources/papers/README.md

@@ -4,15 +4,30 @@
 
 | Resource | Link | Pashto Evidence | Primary Use |
 |---|---|---|---|
+| Benchmark Pashto Handwritten Character Dataset and Pashto Object Character Recognition (OCR) Using Deep Neural Network with Rule Activation Function | [openalex](https://doi.org/10.1155/2021/6669672) | [Matched by explicit Pashto marker in title from OpenAlex works search. (`pashto`)](https://doi.org/10.1155/2021/6669672) | Pashto handwritten OCR benchmark and methodology reference |
 | Benchmarking Whisper for Low-Resource Speech Recognition: An N-Shot Evaluation on Pashto, Punjabi, and Urdu | [other](https://www.semanticscholar.org/paper/13104eddc785756132a19242ac7e74442b145693) | [Matched by Semantic Scholar query: pashto. (`pashto`)](https://www.semanticscholar.org/paper/13104eddc785756132a19242ac7e74442b145693) | Pashto research reference for methods and benchmarking |
+| Database development and automatic speech recognition of isolated Pashto spoken digits using MFCC and K-NN | [openalex](https://doi.org/10.1007/s10772-014-9267-z) | [Matched by explicit Pashto marker in title from OpenAlex works search. (`pashto`)](https://doi.org/10.1007/s10772-014-9267-z) | Pashto ASR baseline method reference for digit recognition |
 | Deep Learning-Based Detection of One and Two-Column Textual Blocks in Camera-Captured Pashto Documents Images | [other](https://www.semanticscholar.org/paper/8c9d2628e23d5c27edc656071f11f0e78124d182) | [Matched by Semantic Scholar query: pashto. (`pashto`)](https://www.semanticscholar.org/paper/8c9d2628e23d5c27edc656071f11f0e78124d182) | Pashto research reference for methods and benchmarking |
+| Development of a New Image-to-text Conversion System for Pashto, Farsi and Traditional Chinese | [arxiv](http://arxiv.org/abs/2005.08650v1) | [Matched by Pashto marker in paper title from arXiv query results. (`pashto`)](http://arxiv.org/abs/2005.08650v1) | Pashto OCR method reference |
 | Enhancing Pashto NER Using Machine-Labeled Data and Transformer-Based Models | [other](https://www.semanticscholar.org/paper/be851ecf9197ef9bb8bf764abf4db0dda95cd9da) | [Matched by explicit Pashto marker in paper title from Semantic Scholar search. (`pashto`)](https://www.semanticscholar.org/paper/be851ecf9197ef9bb8bf764abf4db0dda95cd9da) | Pashto research reference for methods and benchmarking |
 | Enhancing Pashto Text Classification using Language Processing Techniques for Single And Multi-Label Analysis | [arxiv](http://arxiv.org/abs/2305.03201v1) | [Matched by arXiv query: all:pashto. (`pashto`)](http://arxiv.org/abs/2305.03201v1) | Pashto research reference for methods and benchmarking |
+| From Scarcity to Scale: A Release-Level Analysis of the Pashto Common Voice Dataset | [arxiv](http://arxiv.org/abs/2602.14062v1) | [Matched by Pashto marker in paper title from arXiv query results. (`pashto`)](http://arxiv.org/abs/2602.14062v1) | ASR data quality and release trend reference |
 | KNN and ANN-based Recognition of Handwritten Pashto Letters using Zoning Features | [arxiv](http://arxiv.org/abs/1904.03391v2) | [Matched by arXiv query: all:pashto. (`pashto`)](http://arxiv.org/abs/1904.03391v2) | Pashto research reference for methods and benchmarking |
+| KPTI: Katib's Pashto Text Imagebase and Deep Learning Benchmark | [openalex](https://doi.org/10.1109/icfhr.2016.0090) | [Matched by explicit Pashto marker in title from OpenAlex works search. (`pashto`)](https://doi.org/10.1109/icfhr.2016.0090) | Pashto OCR dataset and benchmarking reference |
 | Out-of-Vocabulary Pashto Spell Checker using Morphological Operations | [other](https://www.semanticscholar.org/paper/802aae68a6a7fdfb29d51be03fb2b09e29311fa7) | [Matched by Semantic Scholar query: pashto. (`pashto`)](https://www.semanticscholar.org/paper/802aae68a6a7fdfb29d51be03fb2b09e29311fa7) | Pashto research reference for methods and benchmarking |
+| Pashto isolated digits recognition using deep convolutional neural network | [openalex](https://doi.org/10.1016/j.heliyon.2020.e03372) | [Matched by explicit Pashto marker in title from OpenAlex works search. (`pashto`)](https://doi.org/10.1016/j.heliyon.2020.e03372) | Pashto speech recognition research reference |
+| Pashto offensive language detection: a benchmark dataset and monolingual Pashto BERT | [openalex](https://doi.org/10.7717/peerj-cs.1617) | [Matched by explicit Pashto marker in title from OpenAlex works search. (`pashto`)](https://doi.org/10.7717/peerj-cs.1617) | Pashto NLP toxicity detection benchmark and model reference |
 | Pashto Shallow Parsing: A Deep Learning Approach | [other](https://www.semanticscholar.org/paper/be36455bb4eca60accb3e6866f345132f0dac1e5) | [Matched by Semantic Scholar query: pashto. (`pashto`)](https://www.semanticscholar.org/paper/be36455bb4eca60accb3e6866f345132f0dac1e5) | Pashto research reference for methods and benchmarking |
+| PHTI: Pashto Handwritten Text Imagebase for Deep Learning Applications | [openalex](https://doi.org/10.1109/access.2022.3216881) | [Matched by explicit Pashto marker in title from OpenAlex works search. (`pashto`)](https://doi.org/10.1109/access.2022.3216881) | Pashto OCR dataset and benchmark reference |
+| Pioneer dataset and recognition of Handwritten Pashto characters using Convolution Neural Networks | [openalex](https://doi.org/10.1177/0020294020964826) | [Matched by explicit Pashto marker in title from OpenAlex works search. (`pashto`)](https://doi.org/10.1177/0020294020964826) | Pashto handwritten character recognition reference |
 | POS tagging of low-resource Pashto language: annotated corpus and BERT-based model | [other](https://www.semanticscholar.org/paper/1b2d5c896fec735483e8c8fb0a75e13125e08769) | [Matched by Semantic Scholar query: pashto. (`pashto`)](https://www.semanticscholar.org/paper/1b2d5c896fec735483e8c8fb0a75e13125e08769) | Pashto research reference for methods and benchmarking |
 | PsOCR: Benchmarking Large Multimodal Models for Optical Character Recognition in Low-resource Pashto Language | [other](https://www.semanticscholar.org/paper/d2743c0dcdbc65f5b46fcec2f0ba7cb379c4134f) | [Paper title explicitly references low-resource Pashto language OCR benchmarking. (`Pashto`, `OCR`)](https://www.semanticscholar.org/paper/d2743c0dcdbc65f5b46fcec2f0ba7cb379c4134f) | Pashto OCR research baseline and evaluation reference |
+| Recognition of Pashto Handwritten Characters Based on Deep Learning | [openalex](https://doi.org/10.3390/s20205884) | [Matched by explicit Pashto marker in title from OpenAlex works search. (`pashto`)](https://doi.org/10.3390/s20205884) | Pashto OCR model reference for handwritten character recognition |
+| Recognizable units in Pashto language for OCR | [openalex](https://doi.org/10.1109/icdar.2015.7333963) | [Matched by explicit Pashto marker in title from OpenAlex works search. (`pashto`)](https://doi.org/10.1109/icdar.2015.7333963) | Pashto OCR preprocessing and unit-design reference |
+| Scale and rotation invariant OCR for Pashto cursive script using MDLSTM network | [openalex](https://doi.org/10.1109/icdar.2015.7333931) | [Matched by explicit Pashto marker in title from OpenAlex works search. (`pashto`)](https://doi.org/10.1109/icdar.2015.7333931) | Pashto OCR model architecture reference |
+| Shape analysis of Pashto script and creation of image database for OCR | [openalex](https://doi.org/10.1109/icet.2009.5353160) | [Matched by explicit Pashto marker in title from OpenAlex works search. (`pashto`)](https://doi.org/10.1109/icet.2009.5353160) | Pashto OCR dataset design and feature reference |
+| Speech translation for low-resource languages: the case of Pashto | [openalex](https://doi.org/10.21437/interspeech.2005-723) | [Matched by explicit Pashto marker in title from OpenAlex works search. (`pashto`)](https://doi.org/10.21437/interspeech.2005-723) | Pashto speech translation and low-resource MT reference |
+| Tuning Traditional Language Processing Approaches for Pashto Text Classification | [arxiv](http://arxiv.org/abs/2305.03737v1) | [Matched by Pashto marker in paper title from arXiv query results. (`pashto`)](http://arxiv.org/abs/2305.03737v1) | Pashto text classification method reference |
 
 ## Maintenance
 - Source of truth: [../catalog/resources.json](../catalog/resources.json)

resources/projects/README.md

@@ -5,6 +5,7 @@
 | Resource | Link | Pashto Evidence | Primary Use |
 |---|---|---|---|
 | afaqalinagra/PASHTO-ASR-MODEL | [huggingface](https://huggingface.co/spaces/afaqalinagra/PASHTO-ASR-MODEL) | [Matched by Pashto keyword in Hugging Face Spaces search. (`pashto`)](https://huggingface.co/spaces/afaqalinagra/PASHTO-ASR-MODEL) | Interactive Pashto demo and quick qualitative validation |
+| Aizazayyubi/pashto_asr | [huggingface](https://huggingface.co/spaces/Aizazayyubi/pashto_asr) | [Matched by Pashto keyword in Hugging Face Spaces search. (`pashto`)](https://huggingface.co/spaces/Aizazayyubi/pashto_asr) | Interactive Pashto ASR demo for qualitative evaluation |
 | Fazlullahmamond/Pashto-Typing | [github](https://github.com/Fazlullahmamond/Pashto-Typing) | [Repository metadata (name/description/topics) includes Pashto markers. (`pashto`)](https://github.com/Fazlullahmamond/Pashto-Typing) | Interactive Pashto demo and quick qualitative validation |
 | ihanif/wav2vec-pashto-asr | [huggingface](https://huggingface.co/spaces/ihanif/wav2vec-pashto-asr) | [Matched by Pashto keyword in Hugging Face Spaces search. (`pashto`)](https://huggingface.co/spaces/ihanif/wav2vec-pashto-asr) | Interactive Pashto demo and quick qualitative validation |
 | ihanif/wav2vec2-bert-pashto-asr | [huggingface](https://huggingface.co/spaces/ihanif/wav2vec2-bert-pashto-asr) | [Matched by Pashto keyword in Hugging Face Spaces search. (`pashto`)](https://huggingface.co/spaces/ihanif/wav2vec2-bert-pashto-asr) | Interactive Pashto demo and quick qualitative validation |

resources/schema/resource.schema.json

@@ -1,6 +1,6 @@
 {
   "$schema": "https://json-schema.org/draft/2020-12/schema",
-  "$id": "https://musawer1214.github.io/Pukhto_Pashto/resources/schema/resource.schema.json",
+  "$id": "https://musawer1214.github.io/pashto-language-resources/resources/schema/resource.schema.json",
   "title": "Pashto Resource Catalog",
   "type": "object",
   "additionalProperties": false,
@@ -74,7 +74,13 @@
             "mozilla",
             "kaggle",
             "github",
+            "gitlab",
             "arxiv",
+            "openalex",
+            "crossref",
+            "zenodo",
+            "dataverse",
+            "datacite",
             "meta",
             "other"
           ]
@@ -142,3 +148,4 @@
     }
   }
 }
+

scripts/README.md

@@ -7,7 +7,7 @@ Automation scripts for quality checks, resource catalog validation, and search i
 - `check_links.py`: ensure markdown links are clickable (optional online reachability check).
 - `validate_resource_catalog.py`: validate `resources/catalog/resources.json`.
 - `generate_resource_views.py`: generate `resources/*/README.md`, `resources/README.md`, and `docs/search/resources.json` from the catalog.
-- `sync_resources.py`: collect new candidate Pashto resources from Kaggle, Hugging Face (datasets/models/spaces), GitHub repositories, and paper endpoints into `resources/catalog/pending_candidates.json`.
+- `sync_resources.py`: collect new candidate Pashto resources from Kaggle, Hugging Face (datasets/models/spaces), GitHub, GitLab, OpenAlex, Crossref, Zenodo, Dataverse, DataCite, arXiv, and Semantic Scholar into `resources/catalog/pending_candidates.json`.
 - `run_resource_cycle.py`: run the full repeatable resource cycle with one command.
 
 ## Usage

scripts/sync_resources.py

@@ -21,6 +21,7 @@ import urllib.request
 import xml.etree.ElementTree as ET
 from datetime import datetime, timezone
 from email.utils import parsedate_to_datetime
+from html import unescape
 from http.client import IncompleteRead
 from pathlib import Path
 from typing import Any
@@ -74,6 +75,20 @@ def _is_low_signal_name(value: str) -> bool:
     return bool(LOW_SIGNAL_RE.search(value or ""))
 
 
+def _strip_html(value: str) -> str:
+    text = re.sub(r"<[^>]+>", " ", value or "")
+    text = unescape(text)
+    return re.sub(r"\s+", " ", text).strip()
+
+
+def _classify_repo_category(name_blob: str) -> str:
+    lowered = (name_blob or "").casefold()
+    code_tokens = ("toolkit", "library", "nlp", "asr", "tts", "ocr", "api", "code", "cli", "sdk")
+    if any(token in lowered for token in code_tokens):
+        return "code"
+    return "project"
+
+
 def _parse_retry_after_seconds(retry_after: str | None) -> float | None:
     if not retry_after:
         return None
@@ -440,10 +455,8 @@ def fetch_github_pashto_repos(limit: int) -> list[dict[str, Any]]:
             continue
 
         html_url = item["html_url"]
-        category = "project"
         topics = item.get("topics") or []
-        if any(token in name_blob for token in ("toolkit", "library", "nlp", "asr", "tts", "ocr", "api", "code")):
-            category = "code"
+        category = _classify_repo_category(name_blob)
 
         rid = f"candidate-gh-{category}-{_slug(full_name)}"
         description = (item.get("description") or "").strip()
@@ -467,6 +480,433 @@ def fetch_github_pashto_repos(limit: int) -> list[dict[str, Any]]:
     return out
 
 
+def fetch_gitlab_pashto_projects(limit: int) -> list[dict[str, Any]]:
+    combined: dict[str, dict[str, Any]] = {}
+    errors: list[str] = []
+    for term in PASHTO_QUERY_TERMS:
+        query = urllib.parse.urlencode(
+            {
+                "search": term,
+                "simple": "true",
+                "order_by": "star_count",
+                "sort": "desc",
+                "per_page": str(limit),
+            }
+        )
+        url = f"https://gitlab.com/api/v4/projects?{query}"
+        try:
+            payload = _fetch_json(url, timeout=30.0, source_name="gitlab-projects")
+        except Exception as exc:  # noqa: BLE001
+            errors.append(f"{term}: {exc}")
+            continue
+        for item in payload:
+            full_name = (item.get("path_with_namespace") or item.get("name_with_namespace") or "").strip()
+            if not full_name:
+                continue
+            combined[full_name] = item
+
+    if not combined and errors:
+        raise RuntimeError("; ".join(errors))
+
+    out: list[dict[str, Any]] = []
+    sorted_items = sorted(
+        combined.items(),
+        key=lambda kv: kv[1].get("star_count") or 0,
+        reverse=True,
+    )
+    for full_name, item in sorted_items:
+        web_url = (item.get("web_url") or "").strip()
+        if not web_url:
+            continue
+
+        description = (item.get("description") or "").strip()
+        topics = item.get("topics") or []
+        if not isinstance(topics, list):
+            topics = []
+        topics = [str(topic).strip() for topic in topics if str(topic).strip()]
+        name_blob = " ".join([full_name, item.get("name") or "", description, " ".join(topics)])
+        if not _is_pashto_centric(name_blob):
+            continue
+        if _is_low_signal_name(full_name):
+            continue
+
+        category = _classify_repo_category(name_blob)
+        rid = f"candidate-gitlab-{category}-{_slug(full_name)}"
+        summary = description or "Candidate Pashto-related GitLab repository."
+        out.append(
+            _candidate(
+                rid=rid,
+                title=full_name,
+                url=web_url,
+                category=category,
+                source="gitlab",
+                summary=summary[:240] if summary else "Candidate Pashto-related GitLab repository.",
+                evidence_text="Project metadata (name/description/topics) includes Pashto markers.",
+                evidence_url=web_url,
+                markers=["pashto"],
+                tags=["pashto", "candidate", category, "gitlab", *(topics[:3])],
+            )
+        )
+        if len(out) >= limit:
+            break
+    return out
+
+
+def fetch_openalex_papers(limit: int) -> list[dict[str, Any]]:
+    combined: dict[str, dict[str, Any]] = {}
+    errors: list[str] = []
+    for term in PASHTO_QUERY_TERMS:
+        query = urllib.parse.urlencode({"search": term, "per-page": str(limit)})
+        url = f"https://api.openalex.org/works?{query}"
+        try:
+            payload = _fetch_json(url, timeout=30.0, source_name="openalex")
+        except Exception as exc:  # noqa: BLE001
+            errors.append(f"{term}: {exc}")
+            continue
+        for item in payload.get("results", []):
+            work_id = (item.get("id") or "").strip()
+            if not work_id:
+                continue
+            combined[work_id] = item
+
+    if not combined and errors:
+        raise RuntimeError("; ".join(errors))
+
+    out: list[dict[str, Any]] = []
+    for item in combined.values():
+        title = (item.get("display_name") or "").strip()
+        if not title:
+            continue
+        if not _is_pashto_centric(title):
+            continue
+        if _is_low_signal_name(title):
+            continue
+
+        doi = (item.get("doi") or "").strip()
+        if doi and not doi.startswith("http"):
+            doi = f"https://doi.org/{doi}"
+        primary = item.get("primary_location") or {}
+        landing = (primary.get("landing_page_url") or "").strip()
+        paper_url = doi or landing or (item.get("id") or "").strip()
+        if not paper_url:
+            continue
+
+        rid = f"candidate-openalex-{_slug(title)}"
+        out.append(
+            _candidate(
+                rid=rid,
+                title=title,
+                url=paper_url,
+                category="paper",
+                source="openalex",
+                summary="Candidate paper returned from OpenAlex works search for Pashto.",
+                evidence_text="Matched by explicit Pashto marker in title from OpenAlex works search.",
+                evidence_url=paper_url,
+                markers=["pashto"],
+                tags=["pashto", "candidate", "paper", "openalex"],
+            )
+        )
+        if len(out) >= limit:
+            break
+    return out
+
+
+def fetch_crossref_papers(limit: int) -> list[dict[str, Any]]:
+    combined: dict[str, dict[str, Any]] = {}
+    errors: list[str] = []
+    for term in PASHTO_QUERY_TERMS:
+        query = urllib.parse.urlencode({"query.title": term, "rows": str(limit)})
+        url = f"https://api.crossref.org/works?{query}"
+        try:
+            payload = _fetch_json(url, timeout=30.0, source_name="crossref")
+        except Exception as exc:  # noqa: BLE001
+            errors.append(f"{term}: {exc}")
+            continue
+        for item in payload.get("message", {}).get("items", []):
+            doi = (item.get("DOI") or "").strip()
+            title_list = item.get("title") or []
+            title = (title_list[0] if isinstance(title_list, list) and title_list else "").strip()
+            key = doi or title
+            if not key:
+                continue
+            combined[key] = item
+
+    if not combined and errors:
+        raise RuntimeError("; ".join(errors))
+
+    out: list[dict[str, Any]] = []
+    for item in combined.values():
+        title_list = item.get("title") or []
+        title = (title_list[0] if isinstance(title_list, list) and title_list else "").strip()
+        if not title:
+            continue
+        if not _is_pashto_centric(title):
+            continue
+        if _is_low_signal_name(title):
+            continue
+
+        doi = (item.get("DOI") or "").strip()
+        paper_url = (item.get("URL") or "").strip()
+        if not paper_url and doi:
+            paper_url = f"https://doi.org/{doi}"
+        if not paper_url:
+            continue
+
+        abstract = _strip_html(item.get("abstract") or "")
+        rid = f"candidate-crossref-{_slug(title)}"
+        out.append(
+            _candidate(
+                rid=rid,
+                title=title,
+                url=paper_url,
+                category="paper",
+                source="crossref",
+                summary=(abstract or "Candidate paper returned from Crossref search for Pashto.")[:240],
+                evidence_text="Matched by explicit Pashto marker in title from Crossref search.",
+                evidence_url=paper_url,
+                markers=["pashto"],
+                tags=["pashto", "candidate", "paper", "crossref"],
+            )
+        )
+        if len(out) >= limit:
+            break
+    return out
+
+
+def fetch_zenodo_records(limit: int) -> list[dict[str, Any]]:
+    combined: dict[str, dict[str, Any]] = {}
+    errors: list[str] = []
+    for term in PASHTO_QUERY_TERMS:
+        query = urllib.parse.urlencode({"q": term, "size": str(limit), "sort": "mostrecent"})
+        url = f"https://zenodo.org/api/records/?{query}"
+        try:
+            payload = _fetch_json(url, timeout=30.0, source_name="zenodo")
+        except Exception as exc:  # noqa: BLE001
+            errors.append(f"{term}: {exc}")
+            continue
+        for item in payload.get("hits", {}).get("hits", []):
+            record_id = str(item.get("id") or "").strip()
+            if not record_id:
+                continue
+            combined[record_id] = item
+
+    if not combined and errors:
+        raise RuntimeError("; ".join(errors))
+
+    category_map = {
+        "dataset": "dataset",
+        "software": "code",
+        "publication": "paper",
+        "poster": "project",
+        "presentation": "project",
+    }
+
+    out: list[dict[str, Any]] = []
+    for item in combined.values():
+        metadata = item.get("metadata") or {}
+        title = (metadata.get("title") or "").strip()
+        description = _strip_html(metadata.get("description") or "")
+        if not title:
+            continue
+        if not _is_pashto_centric(title, description):
+            continue
+        if _is_low_signal_name(title):
+            continue
+
+        links = item.get("links") or {}
+        record_url = (links.get("self_html") or links.get("doi") or "").strip()
+        if not record_url:
+            doi = (metadata.get("doi") or "").strip()
+            if doi:
+                record_url = f"https://doi.org/{doi}"
+        if not record_url:
+            continue
+
+        rtype = (metadata.get("resource_type") or {}).get("type") or ""
+        category = category_map.get(str(rtype).casefold(), "project")
+        rid = f"candidate-zenodo-{category}-{_slug(title)}"
+        summary = description or "Candidate resource returned from Zenodo search for Pashto."
+        out.append(
+            _candidate(
+                rid=rid,
+                title=title,
+                url=record_url,
+                category=category,
+                source="zenodo",
+                summary=summary[:240],
+                evidence_text="Zenodo metadata includes Pashto markers in title or description.",
+                evidence_url=record_url,
+                markers=["pashto"],
+                tags=["pashto", "candidate", category, "zenodo"],
+            )
+        )
+        if len(out) >= limit:
+            break
+    return out
+
+
+def fetch_dataverse_datasets(limit: int) -> list[dict[str, Any]]:
+    combined: dict[str, dict[str, Any]] = {}
+    errors: list[str] = []
+    base_url = "https://dataverse.harvard.edu"
+    for term in PASHTO_QUERY_TERMS:
+        query = urllib.parse.urlencode(
+            {
+                "q": term,
+                "type": "dataset",
+                "per_page": str(limit),
+                "start": "0",
+            }
+        )
+        url = f"{base_url}/api/search?{query}"
+        try:
+            payload = _fetch_json(url, timeout=30.0, source_name="dataverse")
+        except Exception as exc:  # noqa: BLE001
+            errors.append(f"{term}: {exc}")
+            continue
+        for item in payload.get("data", {}).get("items", []):
+            key = str(item.get("global_id") or item.get("identifier") or item.get("url") or "").strip()
+            if not key:
+                continue
+            combined[key] = item
+
+    if not combined and errors:
+        raise RuntimeError("; ".join(errors))
+
+    out: list[dict[str, Any]] = []
+    for item in combined.values():
+        title = (item.get("name") or "").strip()
+        description = (item.get("description") or "").strip()
+        if not title:
+            continue
+        if not _is_pashto_centric(title, description):
+            continue
+        if _is_low_signal_name(title):
+            continue
+
+        record_url = (item.get("url") or "").strip()
+        if record_url and record_url.startswith("/"):
+            record_url = f"{base_url}{record_url}"
+        if not record_url:
+            global_id = (item.get("global_id") or "").strip()
+            if global_id:
+                escaped_id = urllib.parse.quote(global_id, safe=":/")
+                record_url = f"{base_url}/dataset.xhtml?persistentId={escaped_id}"
+        if not record_url:
+            continue
+
+        rid = f"candidate-dataverse-dataset-{_slug(title)}"
+        out.append(
+            _candidate(
+                rid=rid,
+                title=title,
+                url=record_url,
+                category="dataset",
+                source="dataverse",
+                summary=(description or "Candidate dataset returned from Dataverse search for Pashto.")[:240],
+                evidence_text="Dataverse metadata includes Pashto markers in dataset title or description.",
+                evidence_url=record_url,
+                markers=["pashto"],
+                tags=["pashto", "candidate", "dataset", "dataverse"],
+            )
+        )
+        if len(out) >= limit:
+            break
+    return out
+
+
+def fetch_datacite_records(limit: int) -> list[dict[str, Any]]:
+    combined: dict[str, dict[str, Any]] = {}
+    errors: list[str] = []
+    for term in PASHTO_QUERY_TERMS:
+        query = urllib.parse.urlencode(
+            {
+                "query": term,
+                "page[size]": str(limit),
+            }
+        )
+        url = f"https://api.datacite.org/dois?{query}"
+        try:
+            payload = _fetch_json(url, timeout=30.0, source_name="datacite")
+        except Exception as exc:  # noqa: BLE001
+            errors.append(f"{term}: {exc}")
+            continue
+        for item in payload.get("data", []):
+            record_id = (item.get("id") or "").strip()
+            if not record_id:
+                continue
+            combined[record_id] = item
+
+    if not combined and errors:
+        raise RuntimeError("; ".join(errors))
+
+    dataset_types = {"dataset", "collection"}
+    software_types = {"software"}
+    paper_types = {"journalarticle", "conferencepaper", "preprint", "text"}
+
+    out: list[dict[str, Any]] = []
+    for item in combined.values():
+        attributes = item.get("attributes") or {}
+        titles = attributes.get("titles") or []
+        title = ""
+        if isinstance(titles, list) and titles:
+            first = titles[0] or {}
+            if isinstance(first, dict):
+                title = (first.get("title") or "").strip()
+        if not title:
+            continue
+        description_items = attributes.get("descriptions") or []
+        descriptions: list[str] = []
+        if isinstance(description_items, list):
+            for block in description_items:
+                if isinstance(block, dict):
+                    value = (block.get("description") or "").strip()
+                    if value:
+                        descriptions.append(_strip_html(value))
+        description_blob = " ".join(descriptions).strip()
+        if not _is_pashto_centric(title, description_blob):
+            continue
+        if _is_low_signal_name(title):
+            continue
+
+        doi = (attributes.get("doi") or item.get("id") or "").strip()
+        record_url = (attributes.get("url") or "").strip()
+        if not record_url and doi:
+            record_url = f"https://doi.org/{doi}"
+        if not record_url:
+            continue
+
+        general_type = str((attributes.get("types") or {}).get("resourceTypeGeneral") or "").casefold()
+        if general_type in dataset_types:
+            category = "dataset"
+        elif general_type in software_types:
+            category = "code"
+        elif general_type in paper_types:
+            category = "paper"
+        else:
+            category = "project"
+
+        rid = f"candidate-datacite-{category}-{_slug(title)}"
+        summary = description_blob or "Candidate record returned from DataCite DOI search for Pashto."
+        out.append(
+            _candidate(
+                rid=rid,
+                title=title,
+                url=record_url,
+                category=category,
+                source="datacite",
+                summary=summary[:240],
+                evidence_text="DataCite metadata includes Pashto markers in title or description.",
+                evidence_url=record_url,
+                markers=["pashto"],
+                tags=["pashto", "candidate", category, "datacite"],
+            )
+        )
+        if len(out) >= limit:
+            break
+    return out
+
+
 def fetch_arxiv(limit: int) -> list[dict[str, Any]]:
     roots: list[ET.Element] = []
     errors: list[str] = []
@@ -651,6 +1091,12 @@ def main() -> int:
         ("huggingface-models", lambda: fetch_huggingface("models", args.limit)),
         ("huggingface-spaces", lambda: fetch_huggingface_spaces(args.limit)),
         ("github-repositories", lambda: fetch_github_pashto_repos(args.limit)),
+        ("gitlab-projects", lambda: fetch_gitlab_pashto_projects(args.limit)),
+        ("openalex", lambda: fetch_openalex_papers(args.limit)),
+        ("crossref", lambda: fetch_crossref_papers(args.limit)),
+        ("zenodo", lambda: fetch_zenodo_records(args.limit)),
+        ("dataverse", lambda: fetch_dataverse_datasets(args.limit)),
+        ("datacite", lambda: fetch_datacite_records(args.limit)),
         ("arxiv", lambda: fetch_arxiv(args.limit)),
         ("semantic-scholar", lambda: fetch_semantic_scholar(args.limit)),
     ]

scripts/validate_resource_catalog.py

@@ -17,7 +17,21 @@ from urllib.parse import urlparse
 
 
 ALLOWED_CATEGORIES = {"dataset", "model", "benchmark", "tool", "paper", "project", "code"}
-ALLOWED_SOURCES = {"huggingface", "mozilla", "kaggle", "github", "arxiv", "meta", "other"}
+ALLOWED_SOURCES = {
+    "huggingface",
+    "mozilla",
+    "kaggle",
+    "github",
+    "gitlab",
+    "arxiv",
+    "openalex",
+    "crossref",
+    "zenodo",
+    "dataverse",
+    "datacite",
+    "meta",
+    "other",
+}
 ALLOWED_STATUS = {"verified", "candidate"}
 RESOURCE_ID_RE = re.compile(r"^[a-z0-9][a-z0-9._-]*$")
 STRICT_PASHTO_CATEGORIES = {"model", "paper", "tool", "code", "project"}