From 14d85392986fe13b2a2ed10f02124ffeba6c2f3d Mon Sep 17 00:00:00 2001
From: Edward Betts <edward@4angle.com>
Date: Sat, 9 Dec 2023 18:42:53 +0000
Subject: [PATCH] Link matching improvements

---
 add_links/match.py | 79 ++++++++++++++++++++++++++++------------------
 1 file changed, 49 insertions(+), 30 deletions(-)

diff --git a/add_links/match.py b/add_links/match.py
index 3ea90a5..41c9ef5 100644
--- a/add_links/match.py
+++ b/add_links/match.py
@@ -23,7 +23,7 @@ trans2[en_dash] = trans2[" "]
 
 patterns = [
     lambda q: re.compile(
-        r"(?<!-)(?:\[\[(?:[^]]+\|)?)?(%s)%s(?:\]\])?"
+        r"(?<!-)\[\[(%s)%s\|(?=.*\]\])"
         % (
             re.escape(q[0]),
             "".join("-?" + (trans2[c] if c in trans2 else re.escape(c)) for c in q[1:]),
@@ -31,10 +31,19 @@ patterns = [
         re.I,
     ),
     lambda q: re.compile(
-        r"(?<!-)\[\[[^|]+\|(%s)%s\]\]" % (re.escape(q[0]), re.escape(q[1:])), re.I
+        r"(?<!-)\[\[(?:(?!File:)(?:[^]]+\|)?)(%s)%s\]\]"
+        % (
+            re.escape(q[0]),
+            "".join("-?" + (trans2[c] if c in trans2 else re.escape(c)) for c in q[1:]),
+        ),
+        re.I,
     ),
     lambda q: re.compile(
-        r"(?<!-)\[\[[^|]+\|(%s)%s(?:\]\])?"
+        r"(?<!-)\[\[(?!File:)[^|]+\|(%s)%s\]\]" % (re.escape(q[0]), re.escape(q[1:])),
+        re.I,
+    ),
+    lambda q: re.compile(
+        r"(?<!-)\[\[(?!File:)[^|]+\|(%s)%s(?:\]\])?"
         % (
             re.escape(q[0]),
             "".join("-?" + (trans2[c] if c in trans2 else re.escape(c)) for c in q[1:]),
@@ -46,7 +55,7 @@ patterns = [
         r"(?<!-)(%s)%s"
         % (
             re.escape(q[0]),
-            "".join((trans[c] if c in trans else re.escape(c)) for c in q[1:]),
+            "".join((trans2[c] if c in trans2 else re.escape(c)) for c in q[1:]),
         ),
         re.I,
     ),
@@ -58,7 +67,9 @@ class NoMatch(Exception):
 
 
 re_cite = re.compile(
-    r"<ref( [^>]*?)?>\s*({{cite.*?}}|\[https?://[^]]*?\])\s*</ref>", re.I | re.S
+    # r"<ref( [^>]*?)?>\s*({{cite.*?}}|\[https?://[^]]*?\])\s*</ref>", re.I | re.S
+    r"<ref( [^>]*?)?>.*</ref>",
+    re.I | re.S,
 )
 
 
@@ -173,8 +184,14 @@ def mk_link_matcher(q: str) -> typing.Callable[[str], re.Match[str] | None]:
     return search_for_link
 
 
-def add_link(m, replacement, text):
-    return m.re.sub(lambda m: "[[%s]]" % replacement, text, count=1)
+def add_link(m: re.Match[str], replacement: str, text: str) -> str:
+    """Add link to text."""
+
+    matched_text = m.group(0)
+    if matched_text.startswith("[[") and matched_text.endswith("|"):
+        return m.re.sub(lambda m: f"[[{replacement}|", text, count=1)
+    else:
+        return m.re.sub(lambda m: f"[[{replacement}]]", text, count=1)
 
 
 def find_link_in_chunk(
@@ -303,30 +320,32 @@ def find_link_and_section(q: str, content: str, linkto: str | None = None):
         if header:
             new_content += header
         for token_type, text in parse_cite(section_text):
-            if token_type == "text" and not replacement:
-                new_text = ""
-                for token_type2, text2 in parse_links(text):
-                    if token_type2 == "link" and not replacement:
-                        link_text = text2[2:-2]
-                        if "|" in link_text:
-                            link_dest, link_text = link_text.split("|", 1)
-                        else:
-                            link_dest = None
-                        m = search_for_link(link_text)
-                        if m:
-                            if link_dest:
-                                found["link_dest"] = link_dest
-                            found["link_text"] = link_text
-                            replacement = match_found(m, q, None)
-                            text2 = add_link(m, replacement, link_text)
-                    new_text += text2
-                if replacement:
-                    text = new_text
-                else:
-                    m = search_for_link(text)
+            if token_type != "text" or replacement:
+                new_content += text
+                continue
+            new_text = ""
+            for token_type2, text2 in parse_links(text):
+                if token_type2 == "link" and not replacement:
+                    link_text = text2[2:-2]
+                    if "|" in link_text:
+                        link_dest, link_text = link_text.split("|", 1)
+                    else:
+                        link_dest = None
+                    m = search_for_link(link_text)
                     if m:
-                        replacement = match_found(m, q, linkto)
-                        text = add_link(m, replacement, text)
+                        if link_dest:
+                            found["link_dest"] = link_dest
+                        found["link_text"] = link_text
+                        replacement = match_found(m, q, None)
+                        text2 = add_link(m, replacement, link_text)
+                new_text += text2
+            if replacement:
+                text = new_text
+            else:
+                m = search_for_link(text)
+                if m:
+                    replacement = match_found(m, q, linkto)
+                    text = add_link(m, replacement, text)
             new_content += text
         if replacement:
             found.update(