File size: 6,838 Bytes
1ef1de7 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 | from __future__ import annotations
import json
import struct
from copy import deepcopy
from pathlib import Path
from ocrtech.cli import main
from ocrtech.language_registry import audit_font_renderability, audit_language_registry
ROOT = Path(__file__).resolve().parents[1]
REGISTRY = ROOT / "languages" / "nepal-v1.json"
def test_nepal_language_registry_passes_core_gates(tmp_path: Path) -> None:
audit = audit_language_registry(REGISTRY, tmp_path / "audit")
assert audit.passed
assert audit.language_count >= 11
assert audit.counted_nepal_language_count >= 10
assert audit.first_target_ids == ["limbu"]
assert audit.primary_scripts["limbu"] == ["Limbu / Sirijonga"]
assert "english" in audit.primary_scripts
assert not audit.errors
assert (tmp_path / "audit" / "language-registry-audit.json").exists()
assert (tmp_path / "audit" / "language-registry-audit.md").exists()
def test_language_registry_rejects_non_limbu_first(tmp_path: Path) -> None:
payload = json.loads(REGISTRY.read_text(encoding="utf-8"))
payload["languages"] = [payload["languages"][1], payload["languages"][0], *payload["languages"][2:]]
registry = tmp_path / "bad-order.json"
registry.write_text(json.dumps(payload, ensure_ascii=False), encoding="utf-8")
audit = audit_language_registry(registry, tmp_path / "audit")
assert not audit.passed
assert "first language entry must have id 'limbu'" in audit.errors
def test_language_registry_rejects_too_few_counted_nepal_languages(tmp_path: Path) -> None:
payload = json.loads(REGISTRY.read_text(encoding="utf-8"))
reduced = deepcopy(payload)
counted = 0
for language in reduced["languages"]:
if language["counts_toward_nepal_language_target"]:
counted += 1
if counted > 3:
language["counts_toward_nepal_language_target"] = False
registry = tmp_path / "too-few.json"
registry.write_text(json.dumps(reduced, ensure_ascii=False), encoding="utf-8")
audit = audit_language_registry(registry, tmp_path / "audit", min_nepal_languages=10)
assert not audit.passed
assert any(error.startswith("registry counts 3 Nepal languages") for error in audit.errors)
def test_language_registry_rejects_english_counting_toward_target(tmp_path: Path) -> None:
payload = json.loads(REGISTRY.read_text(encoding="utf-8"))
for language in payload["languages"]:
if language["id"] == "english":
language["counts_toward_nepal_language_target"] = True
registry = tmp_path / "english-counted.json"
registry.write_text(json.dumps(payload, ensure_ascii=False), encoding="utf-8")
audit = audit_language_registry(registry, tmp_path / "audit")
assert not audit.passed
assert "english must not count toward the Nepal-language target" in audit.errors
def test_cli_audit_language_registry_writes_summary(tmp_path: Path, capsys) -> None:
exit_code = main(["audit-language-registry", str(REGISTRY), "--out", str(tmp_path / "audit")])
captured = capsys.readouterr()
payload = json.loads((tmp_path / "audit" / "language-registry-audit.json").read_text(encoding="utf-8"))
assert exit_code == 0
assert "passed=True" in captured.out
assert "first_targets=limbu" in captured.out
assert payload["counted_nepal_language_count"] >= 10
def test_font_renderability_audit_scores_cmap_coverage(tmp_path: Path) -> None:
font = tmp_path / "single-a.ttf"
_write_minimal_cmap_font(font, 0x0041)
registry = tmp_path / "registry.json"
registry.write_text(
json.dumps(
{
"languages": [
{
"id": "fixture",
"scripts": [
{
"script": "Latin",
"primary": True,
"unicode_ranges": ["U+0041"],
}
],
"font_candidates": [
{"name": "Fixture", "path": str(font), "status": "pending_render_audit"}
],
}
]
}
),
encoding="utf-8",
)
audit = audit_font_renderability(registry, tmp_path / "audit")
assert audit.passed
assert audit.script_count == 1
assert audit.candidates[0].coverage_ratio == 1.0
assert (tmp_path / "audit" / "font-renderability-audit.json").exists()
assert (tmp_path / "audit" / "font-renderability-audit.md").exists()
def test_cli_font_renderability_audit_fails_missing_coverage(tmp_path: Path, capsys) -> None:
font = tmp_path / "single-a.ttf"
_write_minimal_cmap_font(font, 0x0041)
registry = tmp_path / "registry.json"
registry.write_text(
json.dumps(
{
"languages": [
{
"id": "fixture",
"scripts": [
{
"script": "Latin",
"primary": True,
"unicode_ranges": ["U+0042"],
}
],
"font_candidates": [
{"name": "Fixture", "path": str(font), "status": "pending_render_audit"}
],
}
]
}
),
encoding="utf-8",
)
exit_code = main(["audit-font-renderability", str(registry), "--out", str(tmp_path / "audit")])
captured = capsys.readouterr()
payload = json.loads((tmp_path / "audit" / "font-renderability-audit.json").read_text(encoding="utf-8"))
assert exit_code == 1
assert "passed=False" in captured.out
assert payload["errors"] == ["fixture:Latin: no font candidate covers at least 95.00% of required codepoints"]
assert payload["candidates"][0]["missing_sample"] == ["U+0042"]
def _write_minimal_cmap_font(path: Path, codepoint: int) -> None:
seg_count = 2
end_codes = [codepoint, 0xFFFF]
start_codes = [codepoint, 0xFFFF]
id_deltas = [1, 1]
id_range_offsets = [0, 0]
subtable = b"".join(
[
struct.pack(">HHHHHHH", 4, 32, 0, seg_count * 2, 0, 0, 0),
struct.pack(">HH", *end_codes),
struct.pack(">H", 0),
struct.pack(">HH", *start_codes),
struct.pack(">hh", *id_deltas),
struct.pack(">HH", *id_range_offsets),
]
)
cmap = struct.pack(">HHHHL", 0, 1, 3, 1, 12) + subtable
offset_table = struct.pack(">LHHHH", 0x00010000, 1, 0, 0, 0)
table_record = b"cmap" + struct.pack(">LLL", 0, len(offset_table) + 16, len(cmap))
path.write_bytes(offset_table + table_record + cmap)
|