Bug 693639: some convenience functions.

Added primarily for use by SumatraPDF.

Thanks to zeniko.

3 files changed, 76 insertions, 0 deletions

diff --git a/pdf/mupdf.h b/pdf/mupdf.h
index 16a3053d..83710b84 100644
--- a/pdf/mupdf.h
+++ b/pdf/mupdf.h
@@ -27,6 +27,8 @@ pdf_obj *pdf_new_matrix(fz_context *ctx, const fz_matrix *mtx);
 pdf_obj *pdf_copy_array(fz_context *ctx, pdf_obj *array);
 pdf_obj *pdf_copy_dict(fz_context *ctx, pdf_obj *dict);
 
+pdf_obj *pdf_new_obj_from_str(fz_context *ctx, const char *src);
+
 pdf_obj *pdf_keep_obj(pdf_obj *obj);
 void pdf_drop_obj(pdf_obj *obj);
 
@@ -64,6 +66,7 @@ int pdf_array_len(pdf_obj *array);
 pdf_obj *pdf_array_get(pdf_obj *array, int i);
 void pdf_array_put(pdf_obj *array, int i, pdf_obj *obj);
 void pdf_array_push(pdf_obj *array, pdf_obj *obj);
+void pdf_array_push_drop(pdf_obj *array, pdf_obj *obj);
 void pdf_array_insert(pdf_obj *array, pdf_obj *obj);
 int pdf_array_contains(pdf_obj *array, pdf_obj *obj);
 
@@ -94,6 +97,7 @@ char *pdf_to_utf8(pdf_document *xref, pdf_obj *src);
 unsigned short *pdf_to_ucs2(pdf_document *xref, pdf_obj *src); /* sumatrapdf */
 pdf_obj *pdf_to_utf8_name(pdf_document *xref, pdf_obj *src);
 char *pdf_from_ucs2(pdf_document *xref, unsigned short *str);
+void pdf_to_ucs2_buf(unsigned short *buffer, pdf_obj *src);
 
 fz_rect *pdf_to_rect(fz_context *ctx, pdf_obj *array, fz_rect *rect);
 fz_matrix *pdf_to_matrix(fz_context *ctx, pdf_obj *array, fz_matrix *mat);
diff --git a/pdf/pdf_object.c b/pdf/pdf_object.c
index a2d9d37b..c3092afe 100644
--- a/pdf/pdf_object.c
+++ b/pdf/pdf_object.c
@@ -544,6 +544,25 @@ pdf_array_push(pdf_obj *obj, pdf_obj *item)
 }
 
 void
+pdf_array_push_drop(pdf_obj *obj, pdf_obj *item)
+{
+	fz_context *ctx = obj->ctx;
+
+	fz_try(ctx)
+	{
+		pdf_array_push(obj, item);
+	}
+	fz_always(ctx)
+	{
+		pdf_drop_obj(item);
+	}
+	fz_catch(ctx)
+	{
+		fz_rethrow(ctx);
+	}
+}
+
+void
 pdf_array_insert(pdf_obj *obj, pdf_obj *item)
 {
 	RESOLVE(obj);
@@ -1191,6 +1210,30 @@ pdf_drop_obj(pdf_obj *obj)
 		fz_free(obj->ctx, obj);
 }
 
+pdf_obj *pdf_new_obj_from_str(fz_context *ctx, const char *src)
+{
+	pdf_obj *result;
+	pdf_lexbuf lexbuf;
+	fz_stream *stream = fz_open_memory(ctx, (unsigned char *)src, strlen(src));
+
+	pdf_lexbuf_init(ctx, &lexbuf, PDF_LEXBUF_SMALL);
+	fz_try(ctx)
+	{
+		result = pdf_parse_stm_obj(NULL, stream, &lexbuf);
+	}
+	fz_always(ctx)
+	{
+		pdf_lexbuf_fin(&lexbuf);
+		fz_close(stream);
+	}
+	fz_catch(ctx)
+	{
+		return NULL;
+	}
+
+	return result;
+}
+
 /* Pretty printing objects */
 
 struct fmt
diff --git a/pdf/pdf_parse.c b/pdf/pdf_parse.c
index 365c9ffd..373ad0e8 100644
--- a/pdf/pdf_parse.c
+++ b/pdf/pdf_parse.c
@@ -150,6 +150,35 @@ pdf_to_ucs2(pdf_document *xref, pdf_obj *src)
 	return dst;
 }
 
+/* allow to convert to UCS-2 without the need for an fz_context */
+/* (buffer must be at least (fz_to_str_len(src) + 1) * 2 bytes in size) */
+void
+pdf_to_ucs2_buf(unsigned short *buffer, pdf_obj *src)
+{
+	unsigned char *srcptr = (unsigned char *) pdf_to_str_buf(src);
+	unsigned short *dstptr = buffer;
+	int srclen = pdf_to_str_len(src);
+	int i;
+
+	if (srclen >= 2 && srcptr[0] == 254 && srcptr[1] == 255)
+	{
+		for (i = 2; i + 1 < srclen; i += 2)
+			*dstptr++ = srcptr[i] << 8 | srcptr[i+1];
+	}
+	else if (srclen >= 2 && srcptr[0] == 255 && srcptr[1] == 254)
+	{
+		for (i = 2; i + 1 < srclen; i += 2)
+			*dstptr++ = srcptr[i] | srcptr[i+1] << 8;
+	}
+	else
+	{
+		for (i = 0; i < srclen; i++)
+			*dstptr++ = pdf_doc_encoding[srcptr[i]];
+	}
+
+	*dstptr = '\0';
+}
+
 /* Convert UCS-2 string into PdfDocEncoding for authentication */
 char *
 pdf_from_ucs2(pdf_document *xref, unsigned short *src)