Fix parsing of HTML returned from raw API

Github must have changed something on their end, the HTML seems to have changed a little.
ekalinin · Sep 21, 2023 · daabc77 · daabc77
1 parent 661b5c5
commit daabc77
Showing 1 changed file with 8 additions and 7 deletions.
diff --git a/gh-md-toc b/gh-md-toc
@@ -1,5 +1,6 @@
 #!/usr/bin/env bash
 
+set -x
 #
 # Steps:
 #
@@ -235,23 +236,23 @@ gh_toc_grab() {
                          }
                          modified_href = modified_href res
                     }
-                    print sprintf("%*s", (level-1)*'"$2"', "") "* [" text "](" gh_url  modified_href ")"
+                    print sprintf("%*s", (level)*'"$2"', "") "* [" text "](" gh_url  modified_href ")"
                     '
     if [ `uname -s` == "OS/390" ]; then
         grepcmd="pcregrep -o"
         echoargs=""
         awkscript='{
-                     level = substr($0, length($0), 1)
-                     text = substr($0, match($0, /a>.*<\/h/)+2, RLENGTH-5)
+                     level = substr($0, 3, 1)
+                     text = substr($0, match($0, />[^<]*<span aria-hidden/)+1, RLENGTH-17)
                      href = substr($0, match($0, "href=\"([^\"]+)?\"")+6, RLENGTH-7)
                      '"$common_awk_script"'
                 }'
     else
         grepcmd="grep -Eo"
         echoargs="-e"
         awkscript='{
-                     level = substr($0, length($0), 1)
-                     text = substr($0, match($0, /a>.*<\/h/)+2, RLENGTH-5)
+                     level = substr($0, 3, 1)
+                     text = substr($0, match($0, />[^<]*<span aria-hidden/)+1, RLENGTH-17)
                      href = substr($0, match($0, "href=\"[^\"]+?\"")+6, RLENGTH-7)
                      '"$common_awk_script"'
                 }'
@@ -266,7 +267,7 @@ gh_toc_grab() {
     sed -e ':a' -e 'N' -e '$!ba' -e 's/\n<\/h/<\/h/g' |
 
     # find strings that corresponds to template
-    $grepcmd '<a.*id="user-content-[^"]*".*</h[1-6]' |
+    $grepcmd '<h.*id="user-content-[^"]*".*</h[1-6]' |
 
     # remove code tags
     sed 's/<code>//g' | sed 's/<\/code>//g' |
@@ -275,7 +276,7 @@ gh_toc_grab() {
     sed 's/<g-emoji[^>]*[^<]*<\/g-emoji> //g' |
 
     # now all rows are like:
-    #   <a id="user-content-..." href="..."><span ...></span></a> ... </h1
+    #   <h1 id="user-content-..."><a href="..."> ... <span ...></span></a></h1
     # format result line
     #   * $0 - whole string
     #   * last element of each row: "</hN" where N in (1,2,3,...)