Adjust parsers and prefix mapping

esgen/utils.py

@@ -1,3 +1,4 @@
+import re
 import requests
 
 from esgen.config import N_SEARCH_RESULTS
@@ -26,7 +27,7 @@ def wikidata_api_search(inputs: str = " ", search_type: str = "item") -> list:
         choices = list()
         for item in data:
             try:
-                name = f"{item['display']['label']['value']} ({item['id']}) {item['display']['description']['value']}"
+                name = f"{item['display']['label']['value']} ({item['id']}): {item['display']['description']['value']}"
                 value = f"{item['id']}"
                 # (name: the displayed name of the checkbox button, value: the value to be passed to the function)
                 choices.append((name, value))
@@ -43,28 +44,125 @@ def unify_comment(comment: str) -> str:
     :param comment:
     :return:
     """
-    if comment.startswith(" #"):
-        return comment
-    elif comment.startswith("#"):
-        return " " + comment
-    elif comment.startswith(" "):
-        return " #" + comment
+    return comment.strip().lstrip("#").strip()
+
+
+def prefix_mapping(prefix_mappings, entity_string):
+    """
+
+    :param prefix_mappings:
+    :param entity_string:
+    :return:
+    """
+    if entity_string.startswith("wdt:") or entity_string.startswith("wd:"):
+        return entity_string
+    elif re.match(r"P[0-9]+", entity_string):
+        return f"wdt:{entity_string}"
+    elif re.match(r"Q[0-9]+", entity_string):
+        return f"wd:{entity_string}"
     else:
-        return " # " + comment
+        for abbr, prefix in prefix_mappings.items():
+            entity_string = entity_string.replace(prefix, abbr + ":")
+        return entity_string
 
 
-# TODO: INCOMPLETE
-def prefix_mapping(text: str, prefix: dict, prefix_type: str) -> str:
+# parsers
+def shex_blocks_parser(inputs: str) -> list:
     """
+    assumptions:
+      only has patterns like:
+      ```
+      <a> {
+        ... (no brackets inside)
+      }
+
+      <b> {
+        ... (no brackets inside)
+      }
+      ```
 
-    :param text:
-    :param prefix:
-    :param prefix_type:
+    :param inputs:
     :return:
     """
-    if prefix_type == "property":
-        return "wdt:{}".format(text)
-    elif prefix_type == "class":
-        return "wd:{}".format(text)
+    shex_pattern = r"\{([\S\n\t\v ]*?)\}"
+    shex_blocks = re.findall(shex_pattern, inputs)
+    return shex_blocks
+
+
+def prefix_parser(inputs: str) -> [str]:
+    """
+
+    :param inputs:
+    :return:
+    """
+    inputs = inputs.replace("PREFIX", "")
+    split_index = inputs.index(":")
+    prefix_abbr = inputs[:split_index].strip()
+    prefix_url = inputs[split_index + 1:].strip().lstrip("<").rstrip(">")
+    return prefix_abbr, prefix_url
+
+
+def value_shape_name_parser(inputs: str) -> list:
+    """
+
+    :param inputs:
+    :return:
+    """
+    value_shape_names = []
+
+    lines = inputs.split("\n")
+    for line in lines:
+        # if not line.startswith("start"):
+        #     matches = re.findall(r"@<.+>", line)
+        if line.startswith("<"):
+            matches = re.findall(r"<.+>", line)
+            if matches:
+                matches = [name.lstrip("@<").rstrip(">") for name in matches]
+            value_shape_names += matches
+    return value_shape_names
+
+
+def triple_constraint_parser(inputs: str) -> [str]:
+    """
+    case: wdt:P31 [ wd:Q5] ;
+
+    :param inputs:
+    :return:
+    """
+    inputs = inputs.strip().rstrip(";").strip()
+    inputs = re.sub(r"\s+", " ", inputs)
+    # constraint type
+    constraint_type = "triple"
+    # prop
+    prop = inputs.split()[0]
+    # value
+    if " . " in inputs:
+        value = "."
+    elif "[" and "]" in inputs:  # value sets
+        # value = re.findall(r"[[].*]", inputs)[0]  # Possible nested set
+        value = re.findall(r"\[([^\[\]]*)\]", inputs)[0]
+        value = value.split()  # list of values
+    elif "@<" in inputs:  # value shape
+        value = re.findall(r"<.+>", inputs)[0]
+        value = value[1:-1]
+        constraint_type = "value_shape"
+    else:
+        value = inputs.split()[1]
+    # comment
+    if "#" in inputs:
+        comment = unify_comment(inputs[inputs.index("#"):])
     else:
-        return text
+        comment = ""
+    # cardinality
+    if "?" in inputs:
+        cardinality = "?"
+    elif "*" in inputs:
+        cardinality = "*"
+    elif "+" in inputs:
+        cardinality = "+"
+    elif re.search(r"{.+}", inputs):
+        cardinality = re.findall(r"{.+}", inputs)[0]
+    else:
+        cardinality = ""
+
+    return prop, value, cardinality, comment, constraint_type