Fossil SCM

Tweak [1f20d61a07] to avoid broken UTF-8 sequences at the end of truncated comments.

florian 2026-02-07 15:08 trunk

Commit f661b8de491e40daa44c37eba66f02cca0acfa0cc2a683bc577071412699b1ea

Parent df668553c0f0b47…

2 files changed +12 -1 +5 -8

M src/blob.c

+12 -1

		--- src/blob.c
		+++ src/blob.c
		@@ -744,15 +744,26 @@
744	744	void blob_rewind(Blob *p){
745	745	p->iCursor = 0;
746	746	}
747	747
748	748	/*
749		-** Truncate a blob back to zero length
	749	+** Truncate a blob to the specified length in bytes.
750	750	*/
751	751	void blob_truncate(Blob *p, int sz){
752	752	if( sz>=0 && sz<(int)(p->nUsed) ) p->nUsed = sz;
753	753	}
	754	+
	755	+/*
	756	+** Truncate a blob to the specified length in bytes. If truncation
	757	+** results in an incomplete UTF-8 sequence at the end, remove up
	758	+** to three more bytes back to the last complete UTF-8 sequence.
	759	+*/
	760	+void blob_truncate_utf8(Blob *p, int sz){
	761	+ if( sz>=0 && sz<(int)(p->nUsed) ){
	762	+ p->nUsed = utf8_nearest_codepoint(p->aData,sz);
	763	+ }
	764	+}
754	765
755	766	/*
756	767	** Seek the cursor in a blob to the indicated offset.
757	768	*/
758	769	int blob_seek(Blob *p, int offset, int whence){
759	770

	--- src/blob.c
	+++ src/blob.c
	@@ -744,15 +744,26 @@
744	void blob_rewind(Blob *p){
745	p->iCursor = 0;
746	}
747
748	/*
749	** Truncate a blob back to zero length
750	*/
751	void blob_truncate(Blob *p, int sz){
752	if( sz>=0 && sz<(int)(p->nUsed) ) p->nUsed = sz;
753	}











754
755	/*
756	** Seek the cursor in a blob to the indicated offset.
757	*/
758	int blob_seek(Blob *p, int offset, int whence){
759

	--- src/blob.c
	+++ src/blob.c
	@@ -744,15 +744,26 @@
744	void blob_rewind(Blob *p){
745	p->iCursor = 0;
746	}
747
748	/*
749	** Truncate a blob to the specified length in bytes.
750	*/
751	void blob_truncate(Blob *p, int sz){
752	if( sz>=0 && sz<(int)(p->nUsed) ) p->nUsed = sz;
753	}
754
755	/*
756	** Truncate a blob to the specified length in bytes. If truncation
757	** results in an incomplete UTF-8 sequence at the end, remove up
758	** to three more bytes back to the last complete UTF-8 sequence.
759	*/
760	void blob_truncate_utf8(Blob *p, int sz){
761	if( sz>=0 && sz<(int)(p->nUsed) ){
762	p->nUsed = utf8_nearest_codepoint(p->aData,sz);
763	}
764	}
765
766	/*
767	** Seek the cursor in a blob to the indicated offset.
768	*/
769	int blob_seek(Blob *p, int offset, int whence){
770

M src/timeline.c

+5 -8

		--- src/timeline.c
		+++ src/timeline.c
		@@ -731,12 +731,12 @@
731	731	wiki_convert(&comment, 0, WIKI_INLINE);
732	732	}
733	733	wiki_hyperlink_override(0);
734	734	}else{
735	735	if( mxWikiLen>0 && blob_size(&comment)>mxWikiLen ){
736		- blob_truncate(&comment, mxWikiLen);
737		- blob_appendf(&comment, "...");
	736	+ blob_truncate_utf8(&comment, mxWikiLen);
	737	+ blob_append(&comment, "...", 3);
738	738	}
739	739	wiki_convert(&comment, 0, WIKI_INLINE);
740	740	}
741	741	}else{
742	742	if( bCommentGitStyle ){
		@@ -751,16 +751,13 @@
751	751	}
752	752	}
753	753	z[ii] = 0;
754	754	cgi_printf("%W",z);
755	755	}else if( mxWikiLen>0 && (int)blob_size(&comment)>mxWikiLen ){
756		- Blob truncated;
757		- blob_zero(&truncated);
758		- blob_append(&truncated, blob_buffer(&comment), mxWikiLen);
759		- blob_append(&truncated, "...", 3);
760		- @ %W(blob_str(&truncated))
761		- blob_reset(&truncated);
	756	+ blob_truncate_utf8(&comment, mxWikiLen);
	757	+ blob_append(&comment, "...", 3);
	758	+ @ %W(blob_str(&comment))
762	759	drawDetailEllipsis = 0;
763	760	}else{
764	761	cgi_printf("%W",blob_str(&comment));
765	762	}
766	763	}
767	764

	--- src/timeline.c
	+++ src/timeline.c
	@@ -731,12 +731,12 @@
731	wiki_convert(&comment, 0, WIKI_INLINE);
732	}
733	wiki_hyperlink_override(0);
734	}else{
735	if( mxWikiLen>0 && blob_size(&comment)>mxWikiLen ){
736	blob_truncate(&comment, mxWikiLen);
737	blob_appendf(&comment, "...");
738	}
739	wiki_convert(&comment, 0, WIKI_INLINE);
740	}
741	}else{
742	if( bCommentGitStyle ){
	@@ -751,16 +751,13 @@
751	}
752	}
753	z[ii] = 0;
754	cgi_printf("%W",z);
755	}else if( mxWikiLen>0 && (int)blob_size(&comment)>mxWikiLen ){
756	Blob truncated;
757	blob_zero(&truncated);
758	blob_append(&truncated, blob_buffer(&comment), mxWikiLen);
759	blob_append(&truncated, "...", 3);
760	@ %W(blob_str(&truncated))
761	blob_reset(&truncated);
762	drawDetailEllipsis = 0;
763	}else{
764	cgi_printf("%W",blob_str(&comment));
765	}
766	}
767

	--- src/timeline.c
	+++ src/timeline.c
	@@ -731,12 +731,12 @@
731	wiki_convert(&comment, 0, WIKI_INLINE);
732	}
733	wiki_hyperlink_override(0);
734	}else{
735	if( mxWikiLen>0 && blob_size(&comment)>mxWikiLen ){
736	blob_truncate_utf8(&comment, mxWikiLen);
737	blob_append(&comment, "...", 3);
738	}
739	wiki_convert(&comment, 0, WIKI_INLINE);
740	}
741	}else{
742	if( bCommentGitStyle ){
	@@ -751,16 +751,13 @@
751	}
752	}
753	z[ii] = 0;
754	cgi_printf("%W",z);
755	}else if( mxWikiLen>0 && (int)blob_size(&comment)>mxWikiLen ){
756	blob_truncate_utf8(&comment, mxWikiLen);
757	blob_append(&comment, "...", 3);
758	@ %W(blob_str(&comment))



759	drawDetailEllipsis = 0;
760	}else{
761	cgi_printf("%W",blob_str(&comment));
762	}
763	}
764