From 3cec84786639a0def1e22e8dc15be00f37be1230 Mon Sep 17 00:00:00 2001
From: Mike Gerwitz <mike@mikegerwitz.com>
Date: Sun, 10 Mar 2013 21:51:08 -0400
Subject: [PATCH] Added cref-errlog to post-process cref errors rather than
 priming the hashcache

This is more performant, contains additional logging and will properly output
invalid crefs.
---
 git/commit2html |  2 +-
 msgfmt          | 15 +++++++++++++--
 processor       |  4 ++--
 repo2html       | 35 +++++++++++++++++++++++++++--------
 4 files changed, 43 insertions(+), 13 deletions(-)

diff --git a/git/commit2html b/git/commit2html
index 69a64df..a8b9acc 100755
--- a/git/commit2html
+++ b/git/commit2html
@@ -28,6 +28,6 @@ msgfmt="${msgfmt:-./msgfmt}"
 echo -n '<p>'
 
 git show --pretty=format:%b "$commit" \
-  | "$msgfmt"
+  | "$msgfmt" "$commit"
 
 echo '</p>'
diff --git a/msgfmt b/msgfmt
index 9f1c07a..4a97cee 100755
--- a/msgfmt
+++ b/msgfmt
@@ -20,6 +20,9 @@
 #  along with this program.  If not, see <http://www.gnu.org/licenses/>.
 # #
 
+# optional id (for cref errors)
+id="$1"
+
 # HTML replacements (default)
 lquo='\&ldquo;'
 rquo='\&rdquo;'
@@ -40,9 +43,8 @@ while getopts n opt; do
   esac
 done
 
-
 # format the commit message, stopping at the diff (if any)
-awk -vurl_root="$url_root" '
+awk -vid="$id" -vurl_root="$url_root" -vcref_errlog="$cref_errlog" '
     # replace commit refs with generated URL (allows linking to prior commits
     # without hard-coding the configurable links that could change or be
     # relative to where the content is hosted); this will then be processed as a
@@ -52,6 +54,15 @@ awk -vurl_root="$url_root" '
       # (which will be reflected once we print the line)
       c = "./hashcache " g[1]
       c | getline result
+
+      # if a cref error logfile path was provided, log unknown refs so that they
+      # can be re-processed (if commits are processed in reverse order and the
+      # hashcache is cleared before the run, then this is likely to occur for
+      # every cref)
+      if ( result == "" && cref_errlog && id ) {
+        printf "^" id >cref_errlog
+      }
+
       gsub(/\[cref:.*?\]/, url_root result)
     }
 
diff --git a/processor b/processor
index da94bdb..f94ce5b 100755
--- a/processor
+++ b/processor
@@ -44,8 +44,8 @@ EOH
 
 prevdate=
 lastts=
-firstyear=
-lastyear=
+firstyear=0
+lastyear=0
 lasthash=
 
 # generate index
diff --git a/repo2html b/repo2html
index bfda50c..67e9466 100755
--- a/repo2html
+++ b/repo2html
@@ -78,16 +78,35 @@ msgfmt="${msgfmt:-./msgfmt}"
 # make configuration available to all scripts
 export title desc copyright license msgfmt url_root
 
-# clear and prime the cachefile (TODO: we could maintain cache files from
-# previous runs if we offer a flag that opts out of reprocessing previously
-# processed commits)
-echo "Priming hashcache..." >&2
+# clear the cachefile (TODO: we could maintain cache files from previous runs if
+# we offer a flag that opts out of reprocessing previously processed commits)
+echo "Clearing hashcache..." >&2
 ./hashcache clear
-"$repotype"/list | while read hash _ ts id _; do
-  ./hashcache "$hash" "$( ./outfgen "$ts" "$id" )"
-done
 
-# pass commit list to the HTML and RSS processors
+# set the cref error log, which will allow us to re-process *only* those commits
+# that actually need to be reprocessed; this avoids the need to prime the
+# hashcache, saving cycles
+cref_errlog_first=.cref-errlog
+export cref_errlog="$cref_errlog_first"
+>"$cref_errlog"
+
+# pass commit list to the HTML and RSS processors (cache list to disk for
+# reference and further processing)
+listcache=.list
 "$repotype"/list | tee \
   >( ./rss "$repotype" "$url_root" "$rss_count" > "$path_out/rss.xml" ) \
+  "$listcache" \
   | ./processor "$repotype" "$path_out"
+
+# re-process cref errors (but only once; any errors at this point will be
+# considered to be problem refs)
+export cref_errlog=.cref-bad
+>"$cref_errlog"
+grep -f"$cref_errlog_first" "$listcache" \
+  | ./processor "$repotype" "$path_out" \
+  2> >( sed 's/^/[Reprocessing] /g' >&2 )
+
+# if any invalid crefs remain, then they're bad
+[ -s "$cref_errlog" ] && {
+  echo "warning: bad cref(s); see $cref_errlog" >&2
+}