Merge pull request #9103 from kaushalmodi/encode-urls-in-docs

Encode non-alphanum chars like %, + in URLs in the docs

2 files changed, 8 insertions, 3 deletions

diff --git a/lib/packages/docutils/rstgen.nim b/lib/packages/docutils/rstgen.nim
index 51bb9c757..232da5c93 100644
--- a/lib/packages/docutils/rstgen.nim
+++ b/lib/packages/docutils/rstgen.nim
@@ -424,11 +424,15 @@ proc sortIndex(a: var openArray[IndexEntry]) =
     if h == 1: break
 
 proc escapeLink(s: string): string =
+  ## This proc is mostly copied from uri/encodeUrl except that
+  ## these chars are also left unencoded: '#', '/'.
   result = newStringOfCap(s.len + s.len shr 2)
   for c in items(s):
     case c
-    of 'a'..'z', '_', 'A'..'Z', '0'..'9', '.', '#', ',', '/':
-      result.add c
+    of 'a'..'z', 'A'..'Z', '0'..'9', '-', '.', '_', '~': # same as that in uri/encodeUrl
+      add(result, c)
+    of '#', '/': # example.com/foo/#bar (don't escape the '/' and '#' in such links)
+      add(result, c)
     else:
       add(result, "%")
       add(result, toHex(ord(c), 2))
@@ -444,7 +448,7 @@ proc generateSymbolIndex(symbols: seq[IndexEntry]): string =
     var j = i
     while j < symbols.len and keyword == symbols[j].keyword:
       let
-        url = symbols[j].link #.escapeLink
+        url = symbols[j].link.escapeLink
         text = if symbols[j].linkTitle.len > 0: symbols[j].linkTitle else: url
         desc = if symbols[j].linkDesc.len > 0: symbols[j].linkDesc else: ""
       if desc.len > 0:
diff --git a/lib/pure/uri.nim b/lib/pure/uri.nim
index dd8040928..d296017dd 100644
--- a/lib/pure/uri.nim
+++ b/lib/pure/uri.nim
@@ -60,6 +60,7 @@ proc encodeUrl*(s: string, usePlus=true): string =
   let fromSpace = if usePlus: "+" else: "%20"
   for c in s:
     case c
+    # https://tools.ietf.org/html/rfc3986#section-2.3
     of 'a'..'z', 'A'..'Z', '0'..'9', '-', '.', '_', '~': add(result, c)
     of ' ': add(result, fromSpace)
     else: