shithub: pdffs

Info  •  Files  •  Log  •  Branches

reduce pdfobj recursion on number parsing

Since only the next object needs to be looked ahead (it has to be a number)
we should not recurse further because otherwise parsing integer-only arrays
is going to be rather slow and most likely produce a stack overflow.
Add an explicit "flags" to pdfobj function, with Norecurse set when looking
ahead.

--- a/array.c

+++ b/array.c

@@ -24,7 +24,7 @@

 		}

 		Sungetc(s);

-		if((m = pdfobj(pdf, s)) == nil){

+		if((m = pdfobj(pdf, s, 0)) == nil){

 			noel = 1;

 			continue;

 		}

--- a/dict.c

+++ b/dict.c

@@ -33,7 +33,7 @@

 		}

 		Sungetc(s);

-		if((k = pdfobj(pdf, s)) == nil){

+		if((k = pdfobj(pdf, s, 0)) == nil){

 			nokey = 1;

 			continue;

 		}

@@ -41,7 +41,7 @@

 			werrstr("expected name as a key");

 			goto err;

 		}

-		if((v = pdfobj(pdf, s)) == nil)

+		if((v = pdfobj(pdf, s, 0)) == nil)

 			goto err;

 		if((kv = realloc(o->dict.kv, (o->dict.nkv+1)*sizeof(KeyValue))) == nil)

--- a/eval.c

+++ b/eval.c

@@ -25,7 +25,7 @@

 		werrstr("xref seek failed");

 		goto err;

 	}

-	if((ostm = pdfobj(pdf, pdf->s)) == nil)

+	if((ostm = pdfobj(pdf, pdf->s, 0)) == nil)

 		goto err;

 	first = -1;

 	if((nobj = dictint(ostm, "N")) < 1 || (first = dictint(ostm, "First")) < 0){

@@ -44,7 +44,7 @@

 				werrstr("xref obj seek failed");

 				goto err;

 			}

-			if((o = pdfobj(pdf, s)) == nil)

+			if((o = pdfobj(pdf, s, 0)) == nil)

 				goto err;

 			o = pdfeval(o);

 			break;

@@ -96,7 +96,7 @@

 		werrstr("xref seek failed");

 		return &null;

 	}

-	if((d = pdfobj(o->pdf, o->pdf->s)) == nil){

+	if((d = pdfobj(o->pdf, o->pdf->s, 0)) == nil){

 		werrstr("eval: %r [at %p]", (void*)x->off);

 		return &null;

 	}

--- a/object.c

+++ b/object.c

@@ -52,7 +52,7 @@

 /* General function to parse an object of any type. */

 Object *

-pdfobj(Pdf *pdf, Stream *s)

+pdfobj(Pdf *pdf, Stream *s, int flags)

 {

 	Object *o, *o2;

 	vlong off;

@@ -155,7 +155,8 @@

 		o->num.i = o->num.d;

 		off = Soffset(s); /* seek here if not an indirect object later */

-		if((o2 = pdfobj(pdf, s)) != nil && o2->type == Onum){ /* second object is number too */

+		if((flags & Norecurse) == 0)

+		if((o2 = pdfobj(pdf, s, Norecurse)) != nil && o2->type == Onum){ /* second object is number too */

 			do; while(isws(c = Sgetc(s)));

 			if(c < 0)

 				goto err;

@@ -173,7 +174,7 @@

 				xref.gen = o2->num.i;

 				/* FIXME put into a map */

 				pdfobjfree(o2);

-				if((o2 = pdfobj(pdf, s)) != nil){

+				if((o2 = pdfobj(pdf, s, 0)) != nil){

 					pdfobjfree(o);

 					return o2;

 				}else{

@@ -182,6 +183,7 @@

 				}

 			}

 		}

+		pdfobjfree(o2);

 		/* just a number, go back and return it */

 		o->type = Onum;

--- a/op.c

+++ b/op.c

@@ -1471,7 +1471,7 @@

 			s->buf.off += 1;

 		if(s->buf.off == s->buf.sz)

 			break;

-		o = pdfobj(content->pdf, s);

+		o = pdfobj(content->pdf, s, 0);

 		if(o == nil){

 			werrstr("pagerendercontent: failed to parse op: %r");

 			return 0;

--- a/pdf.c

+++ b/pdf.c

@@ -9,7 +9,7 @@

 	Object *o;

 	int prev;

-	if((o = pdfobj(pdf, pdf->s)) == nil)

+	if((o = pdfobj(pdf, pdf->s, 0)) == nil)

 		goto err;

 	if(o->type != Odict){

 		werrstr("isn't a dictionary");

--- a/pdf.h

+++ b/pdf.h

@@ -206,7 +206,11 @@

 /*

  * Parse an object.

  */

-Object *pdfobj(Pdf *pdf, Stream *s);

+Object *pdfobj(Pdf *pdf, Stream *s, int flags);

+

+enum {

+	Norecurse = 1<<0, /* pdfobj should not recurse (num parsing). */

+};

 /*

  * Deallocate the object and all its children. Refcount is

--- a/xref.c

+++ b/xref.c

@@ -117,7 +117,7 @@

 	int i, ni, nsubsec, subsec;

 	s = nil;

-	if((o = pdfobj(pdf, pdf->s)) == nil){

+	if((o = pdfobj(pdf, pdf->s, 0)) == nil){

 		werrstr("xref stream obj: %r");

 		goto err;

 	}