Exclude binary files from the checks done by the pre-commit hook.

We shouldn't check that binary files are UTF-8 so determine whether the file
is binary by looking at its svn:mime-type and only do the check for the text
files.

git-svn-id: https://svn.wxwidgets.org/svn/wx/wxWidgets/trunk@62045 c3d73ce0-8a6f-49c7-b76d-6d57e0e08775

misc/scripts/svn/hooks/pre-commit

@@ -5,37 +5,56 @@ TXN="$2"
 
 SVNLOOK=/usr/bin/svnlook
 
-svn_cat() {
+svnl() {
-    $SVNLOOK cat "$REPOS" -t "$TXN" $1
+    cmd=$1
+    shift
+    $SVNLOOK $cmd "$REPOS" -t "$TXN" $*
 }
 
-all_changed_files=`$SVNLOOK changed "$REPOS" -t "$TXN" | \
+set -e
-                    grep "^[AU]" | \
-                    sed 's/^....//'`
-
-# notice that breaking all_changed_files into several lines by replacing spaces
-# with new lines only works as long as we don't have any files with spaces in
-# them -- which is the case for now, but if it ever changes we'd probably need
-# to use a shell array for all_changed_files or just rerun svnlook here again
-changed_sources=`echo $all_changed_files | \
-                    sed 's/ /\n/g' | \
-                    egrep "\.(cpp|h|py)$" | \
-                    egrep -v "src/(tiff|regex|jpeg|stc/scintilla)"`
-
 rc=0
 
-set -e
+# exclude all third-party files from consideration, we don't want to do any
+# checks for them
+all_changed_files=`svnl changed | \
+                    grep "^[AU]" | \
+                    sed 's/^....//' | \
+                    egrep -v "src/(tiff|regex|jpeg|stc/scintilla)"`
+
+# analyze the changed files to find all non-binary and all source files
+for f in $all_changed_files; do
+    mimetype=`svnl proplist -v $f |
+                fgrep "svn:mime-type" |
+                sed 's/^ svn:mime-type : //'`
+    case $mimetype in
+        ''|text/*)
+            ;;
+
+        *)
+            continue
+            ;;
+    esac
+
+    changed_text_files="$changed_text_files $f"
+
+    case $f in
+        *.cpp|*.h|*.py)
+            changed_sources="$changed_sources $f"
+            ;;
+    esac
+done
 
 for f in $changed_sources; do
-     if  svn_cat $f | fgrep -q '	'; then
+     if  svnl cat $f | fgrep -q '	'; then
          echo "Please remove TABs from $f before committing." >&2
          rc=1
      fi
 done
 
-for f in $all_changed_files; do
+for f in $changed_text_files; do
-    if ! svn_cat $f | iconv -f utf8 -t WCHAR_T > /dev/null; then
+    if ! svnl cat $f | iconv -f utf8 -t WCHAR_T > /dev/null; then
         echo "File $f doesn't use UTF-8, please convert it before committing." >&2
+        echo "(or set svn:mime-type property correctly if the file is binary)." >&2
         rc=1
     fi
 done