Simplified parsing doc paths

Author: cwhite911

doc/Makefile

@@ -5,8 +5,6 @@ include $(MODULE_TOPDIR)/include/Make/Other.make
 FILES := $(wildcard *.html)
 # So far, we disntinguished user and contributor documentation here by
 # extension. This is no longer possible with Markdown.
-# These are currently maunally added to mam/mkdocs/scripts/hook_list_scripts.py
-# to generate the correct links to edit on GitHub.
 MDFILES := grass_database.md projectionintro.md
 
 default: $(patsubst %,$(HTMLDIR)/%,$(FILES)) $(patsubst %,$(MDDIR)/source/%,$(MDFILES))

man/mkdocs/overrides/partials/actions.html

@@ -1,14 +1,9 @@
 {% if page.edit_url %}
 
     {% set toolname = page.file.src_path.split("/")[-1] | replace(".md", "") %}
-
     {# Default to the normal edit URL #}
-    {% set custom_edit_url = page.edit_url %}
-    {% set ghpath = github_path(toolname, config.extra.scripts_tools) %}
-    <!-- {{ "ghpath: " ~ ghpath }}<br>
-    {{ "toolname: " ~ toolname }}<br> -->
+    {% set ghpath = github_path(toolname, config.extra.source_docs) %}
     {% if ghpath %}
-        {% set custom_edit_url = page.edit_url ~ ghpath %}
         {% set custom_edit_url = page.edit_url | replace(toolname, ghpath ~ "/" ~ toolname) %}
         {% if "content.action.edit" in features %}
             <a href="{{ custom_edit_url }}" title="{{ lang.t('action.edit') }}" class="md-content__button md-icon">

man/mkdocs/scripts/hook_list_scripts.py

@@ -3,92 +3,13 @@
 from jinja2 import Environment
 
 
-def github_path(toolname, scripts_tools):
-    """Finds GitHub path for a tool in GRASS GIS with subtool handling"""
-
-    auto_generated_indexes = [
-        "database",
-        "db",
-        "display",
-        "general",
-        "imagery",
-        "miscellaneous",
-        "postscript",
-        "raster",
-        "raster3d",
-        "temporal",
-        "vector",
-    ]
-
-    # Special cases for documentation pages that don't match the pattern matching scheme
-    special_docs = [
-        {"name": "projectionintro", "path": "doc"},
-        {"name": "grass_database", "path": "doc"},
-        {"name": "databaseintro", "path": "db"},
-    ]
-
-    # Exit early if toolname is empty
-    if not toolname:
+def github_path(toolname: str, source_docs: dict[str, str]):
+    """Return the GitHub path for the given toolname"""
+    try:
+        return source_docs[toolname]
+    except KeyError:
         return None
 
-    # Handle special cases
-    if toolname in [x["name"] for x in special_docs]:
-        return next((x["path"] for x in special_docs if x["name"] == toolname), None)
-
-    # Convert filter() results to a list
-    tool_matches = list(filter(lambda x: toolname in x, scripts_tools))
-
-    # print(f"Tool Matches: {tool_matches}")
-
-    # If there's exactly one match, return it immediately
-    if len(tool_matches) == 1:
-        return tool_matches[0]
-
-    # Prioritize exact matches
-    exact_match = next((x for x in tool_matches if x.endswith(toolname)), None)
-    # print(f"Exact - Tool Matches: {exact_match}")
-    if toolname in auto_generated_indexes:
-        # print(f"Exact Match - Autogenerated: {exact_match}")
-        return None
-    if exact_match:
-        return exact_match
-
-    # Prefer major categories (raster, vector, imagery, temporal, etc.)
-    category_match = next(
-        (x for x in tool_matches if not x.startswith("scripts/")), None
-    )
-    # print(f"Category - Tool Matches: {category_match}")
-    if category_match:
-        return category_match
-
-    # Check for subtools in the same directory (e.g., r.sim -> r.sim.water, r.sim.sediment)
-    subtool_prefix = toolname + "."  # Ensure it matches the full prefix
-    sub_tool_matches = [x for x in tool_matches if x.startswith(subtool_prefix)]
-
-    # print(f"Subtool - Tool Matches: {sub_tool_matches}")
-    # If exactly one subtool match exists, return it
-    if len(sub_tool_matches) == 1:
-        return sub_tool_matches[0]
-
-    # If multiple subtools exist, prioritize based on name length (more specific subtools come first)
-    if sub_tool_matches:
-        sub_tool_matches.sort(key=len)  # Shorter names first
-        return sub_tool_matches[0]  # Return the most specific match
-
-    # Handle special case for "intro" pages
-    # print(f"Intro Doc - Matches: {tool_matches}")
-    if toolname.endswith("intro"):
-        return toolname.replace("intro", "")
-
-    # Special case for gui/wxpython/docs
-    if toolname.startswith("g."):
-        # print(f"GUI Docs: {toolname}")
-        tool_dir = toolname.split(".")[-1]
-        return f"gui/wxpython/{tool_dir}/{toolname}"
-
-    # If nothing else, return the first available match
-    return tool_matches[0] if tool_matches else None
-
 
 def on_env(env: Environment, config, files):
     """Enable loopcontrols extension in Jinja2"""
@@ -103,20 +24,32 @@ def on_config(config):
     store it in MkDocs extra config. These are used to generate
     the correct links to the documentation in GitHub.
     """
-    scripts_dir = Path("source")
-    scripts_tools = []
-    url_pattern = re.compile(
-        r"https://github.com/OSGeo/grass/tree/main/([^ )]+)"
-    )  # Read the mkdocs.yml file
+    source_dir = Path("source")
+
+    # Dict to store the documentation path for each tool in GitHub
+    # the key is the doc name and the value is the path in the GitHub repository
+    source_docs = {}
 
-    for file in scripts_dir.glob("*.md"):
+    # Read the source files and extract the GitHub path from the Available at link at the bottom of the page
+    pattern = re.compile(
+        r"Available at:\s*\[(?P<text>.*?)\]\(https://github\.com/OSGeo/grass/tree/main/(?P<gh_path>.*?)\)"
+    )
+
+    for file in source_dir.glob("*.md"):
         with file.open() as f:
             for line in f:
-                match = url_pattern.search(line)
+                match = pattern.search(line)
                 if match:
-                    toolname = match.group(1).strip()
-                    scripts_tools.append(toolname)
+                    # Extract the entire link text and the GitHub path.
+                    text = match.group("text")
+                    gh_path = match.group("gh_path").strip()
+
+                    # Remove the trailing "source code" from the text to get the tool name.
+                    toolname = re.sub(
+                        r"\s*source\s+code\s*$", "", text, flags=re.IGNORECASE
+                    ).strip()
+                    source_docs[toolname] = gh_path
 
     # Store in MkDocs extra config
-    config["extra"]["scripts_tools"] = scripts_tools
+    config["extra"]["source_docs"] = source_docs
     return config