version bump 1.0.6: performance

- `unsafe` option to `cfb_add` for bulk write (see #2)
- use `lastIndexOf` to save operations in BFP queue

CHANGELOG.md

@@ -4,6 +4,10 @@ This log is intended to keep track of backwards-incompatible changes, including
 but not limited to API changes and file location changes.  Minor behavioral
 changes may not be included if they are not expected to break existing code.
 
+## 1.0.6 (2018-04-09)
+
+* `lastIndexOf` in FAT builder enables larger file parsing at minor cost
+
 ## 1.0.0 (2017-11-05)
 
 * Actually walk mini-fat

README.md

@@ -111,6 +111,7 @@ accept a `name` argument strictly deal with absolute file names:
 
 - `.cfb_new(?opts)` creates a new container object.
 - `.cfb_add(cfb, name, ?content, ?opts)` adds a new file to the `cfb`.
+  Set the option `{unsafe:true}` to skip existence checks (for bulk additions)
 - `.cfb_del(cfb, name)` deletes the specified file
 - `.cfb_mov(cfb, old_name, new_name)` moves the old file to new path and name
 

bits/31_version.js

@@ -1 +1 @@
-exports.version = '1.0.5';
+exports.version = '1.0.6';

bits/43_rbtree.js

@@ -13,8 +13,8 @@ function build_full_paths(FI/*:CFBFileIndex*/, FP/*:Array<string>*/, Paths/*:Arr
 			if(R !== -1 && dad[R] !== R) dad[i] = dad[R];
 		}
 		if(C !== -1 /*NOSTREAM*/) dad[C] = i;
-		if(L !== -1) { dad[L] = dad[i]; q.push(L); }
-		if(R !== -1) { dad[R] = dad[i]; q.push(R); }
+		if(L !== -1) { dad[L] = dad[i]; if(q.lastIndexOf(L) < j) q.push(L); }
+		if(R !== -1) { dad[R] = dad[i]; if(q.lastIndexOf(R) < j) q.push(R); }
 	}
 	for(i=1; i < pl; ++i) if(dad[i] === i) {
 		if(R !== -1 /*NOSTREAM*/ && dad[R] !== R) dad[i] = dad[R];

bits/85_api.js

@@ -5,8 +5,9 @@ function cfb_new(opts/*:?any*/)/*:CFBContainer*/ {
 }
 
 function cfb_add(cfb/*:CFBContainer*/, name/*:string*/, content/*:?RawBytes*/, opts/*:?any*/)/*:CFBEntry*/ {
-	init_cfb(cfb);
-	var file = CFB.find(cfb, name);
+	var unsafe = opts && opts.unsafe;
+	if(!unsafe) init_cfb(cfb);
+	var file = !unsafe && CFB.find(cfb, name);
 	if(!file) {
 		var fpath/*:string*/ = cfb.FullPaths[0];
 		if(name.slice(0, fpath.length) == fpath) fpath = name;
@@ -17,7 +18,7 @@ function cfb_add(cfb/*:CFBContainer*/, name/*:string*/, content/*:?RawBytes*/, o
 		file = ({name: filename(name), type: 2}/*:any*/);
 		cfb.FileIndex.push(file);
 		cfb.FullPaths.push(fpath);
-		CFB.utils.cfb_gc(cfb);
+		if(!unsafe) CFB.utils.cfb_gc(cfb);
 	}
 	/*:: if(!file) throw new Error("unreachable"); */
 	file.content = (content/*:any*/);

cfb.flow.js

@@ -177,7 +177,7 @@ type CFBFiles = {[n:string]:CFBEntry};
 /* [MS-CFB] v20171201 */
 var CFB = (function _CFB(){
 var exports/*:CFBModule*/ = /*::(*/{}/*:: :any)*/;
-exports.version = '1.0.5';
+exports.version = '1.0.6';
 /* [MS-CFB] 2.6.4 */
 function namecmp(l/*:string*/, r/*:string*/)/*:number*/ {
 	var L = l.split("/"), R = r.split("/");
@@ -355,8 +355,8 @@ function build_full_paths(FI/*:CFBFileIndex*/, FP/*:Array<string>*/, Paths/*:Arr
 			if(R !== -1 && dad[R] !== R) dad[i] = dad[R];
 		}
 		if(C !== -1 /*NOSTREAM*/) dad[C] = i;
-		if(L !== -1) { dad[L] = dad[i]; q.push(L); }
-		if(R !== -1) { dad[R] = dad[i]; q.push(R); }
+		if(L !== -1) { dad[L] = dad[i]; if(q.lastIndexOf(L) < j) q.push(L); }
+		if(R !== -1) { dad[R] = dad[i]; if(q.lastIndexOf(R) < j) q.push(R); }
 	}
 	for(i=1; i < pl; ++i) if(dad[i] === i) {
 		if(R !== -1 /*NOSTREAM*/ && dad[R] !== R) dad[i] = dad[R];
@@ -823,8 +823,9 @@ function cfb_new(opts/*:?any*/)/*:CFBContainer*/ {
 }
 
 function cfb_add(cfb/*:CFBContainer*/, name/*:string*/, content/*:?RawBytes*/, opts/*:?any*/)/*:CFBEntry*/ {
-	init_cfb(cfb);
-	var file = CFB.find(cfb, name);
+	var unsafe = opts && opts.unsafe;
+	if(!unsafe) init_cfb(cfb);
+	var file = !unsafe && CFB.find(cfb, name);
 	if(!file) {
 		var fpath/*:string*/ = cfb.FullPaths[0];
 		if(name.slice(0, fpath.length) == fpath) fpath = name;
@@ -835,7 +836,7 @@ function cfb_add(cfb/*:CFBContainer*/, name/*:string*/, content/*:?RawBytes*/, o
 		file = ({name: filename(name), type: 2}/*:any*/);
 		cfb.FileIndex.push(file);
 		cfb.FullPaths.push(fpath);
-		CFB.utils.cfb_gc(cfb);
+		if(!unsafe) CFB.utils.cfb_gc(cfb);
 	}
 	/*:: if(!file) throw new Error("unreachable"); */
 	file.content = (content/*:any*/);

cfb.js

@@ -159,7 +159,7 @@ function new_buf(sz) {
 /* [MS-CFB] v20171201 */
 var CFB = (function _CFB(){
 var exports = {};
-exports.version = '1.0.5';
+exports.version = '1.0.6';
 /* [MS-CFB] 2.6.4 */
 function namecmp(l, r) {
 	var L = l.split("/"), R = r.split("/");
@@ -337,8 +337,8 @@ function build_full_paths(FI, FP, Paths) {
 			if(R !== -1 && dad[R] !== R) dad[i] = dad[R];
 		}
 		if(C !== -1 /*NOSTREAM*/) dad[C] = i;
-		if(L !== -1) { dad[L] = dad[i]; q.push(L); }
-		if(R !== -1) { dad[R] = dad[i]; q.push(R); }
+		if(L !== -1) { dad[L] = dad[i]; if(q.lastIndexOf(L) < j) q.push(L); }
+		if(R !== -1) { dad[R] = dad[i]; if(q.lastIndexOf(R) < j) q.push(R); }
 	}
 	for(i=1; i < pl; ++i) if(dad[i] === i) {
 		if(R !== -1 /*NOSTREAM*/ && dad[R] !== R) dad[i] = dad[R];
@@ -799,8 +799,9 @@ function cfb_new(opts) {
 }
 
 function cfb_add(cfb, name, content, opts) {
-	init_cfb(cfb);
-	var file = CFB.find(cfb, name);
+	var unsafe = opts && opts.unsafe;
+	if(!unsafe) init_cfb(cfb);
+	var file = !unsafe && CFB.find(cfb, name);
 	if(!file) {
 		var fpath = cfb.FullPaths[0];
 		if(name.slice(0, fpath.length) == fpath) fpath = name;
@@ -811,7 +812,7 @@ function cfb_add(cfb, name, content, opts) {
 		file = ({name: filename(name), type: 2});
 		cfb.FileIndex.push(file);
 		cfb.FullPaths.push(fpath);
-		CFB.utils.cfb_gc(cfb);
+		if(!unsafe) CFB.utils.cfb_gc(cfb);
 	}
 file.content = (content);
 	file.size = content ? content.length : 0;

dist/cfb.js

@@ -159,7 +159,7 @@ function new_buf(sz) {
 /* [MS-CFB] v20171201 */
 var CFB = (function _CFB(){
 var exports = {};
-exports.version = '1.0.5';
+exports.version = '1.0.6';
 /* [MS-CFB] 2.6.4 */
 function namecmp(l, r) {
 	var L = l.split("/"), R = r.split("/");
@@ -337,8 +337,8 @@ function build_full_paths(FI, FP, Paths) {
 			if(R !== -1 && dad[R] !== R) dad[i] = dad[R];
 		}
 		if(C !== -1 /*NOSTREAM*/) dad[C] = i;
-		if(L !== -1) { dad[L] = dad[i]; q.push(L); }
-		if(R !== -1) { dad[R] = dad[i]; q.push(R); }
+		if(L !== -1) { dad[L] = dad[i]; if(q.lastIndexOf(L) < j) q.push(L); }
+		if(R !== -1) { dad[R] = dad[i]; if(q.lastIndexOf(R) < j) q.push(R); }
 	}
 	for(i=1; i < pl; ++i) if(dad[i] === i) {
 		if(R !== -1 /*NOSTREAM*/ && dad[R] !== R) dad[i] = dad[R];
@@ -799,8 +799,9 @@ function cfb_new(opts) {
 }
 
 function cfb_add(cfb, name, content, opts) {
-	init_cfb(cfb);
-	var file = CFB.find(cfb, name);
+	var unsafe = opts && opts.unsafe;
+	if(!unsafe) init_cfb(cfb);
+	var file = !unsafe && CFB.find(cfb, name);
 	if(!file) {
 		var fpath = cfb.FullPaths[0];
 		if(name.slice(0, fpath.length) == fpath) fpath = name;
@@ -811,7 +812,7 @@ function cfb_add(cfb, name, content, opts) {
 		file = ({name: filename(name), type: 2});
 		cfb.FileIndex.push(file);
 		cfb.FullPaths.push(fpath);
-		CFB.utils.cfb_gc(cfb);
+		if(!unsafe) CFB.utils.cfb_gc(cfb);
 	}
 file.content = (content);
 	file.size = content ? content.length : 0;

dist/xlscfb.js

@@ -38,7 +38,7 @@ type CFBFiles = {[n:string]:CFBEntry};
 /* [MS-CFB] v20171201 */
 var CFB = (function _CFB(){
 var exports/*:CFBModule*/ = /*::(*/{}/*:: :any)*/;
-exports.version = '1.0.5';
+exports.version = '1.0.6';
 /* [MS-CFB] 2.6.4 */
 function namecmp(l/*:string*/, r/*:string*/)/*:number*/ {
 	var L = l.split("/"), R = r.split("/");
@@ -216,8 +216,8 @@ function build_full_paths(FI/*:CFBFileIndex*/, FP/*:Array<string>*/, Paths/*:Arr
 			if(R !== -1 && dad[R] !== R) dad[i] = dad[R];
 		}
 		if(C !== -1 /*NOSTREAM*/) dad[C] = i;
-		if(L !== -1) { dad[L] = dad[i]; q.push(L); }
-		if(R !== -1) { dad[R] = dad[i]; q.push(R); }
+		if(L !== -1) { dad[L] = dad[i]; if(q.lastIndexOf(L) < j) q.push(L); }
+		if(R !== -1) { dad[R] = dad[i]; if(q.lastIndexOf(R) < j) q.push(R); }
 	}
 	for(i=1; i < pl; ++i) if(dad[i] === i) {
 		if(R !== -1 /*NOSTREAM*/ && dad[R] !== R) dad[i] = dad[R];
@@ -684,8 +684,9 @@ function cfb_new(opts/*:?any*/)/*:CFBContainer*/ {
 }
 
 function cfb_add(cfb/*:CFBContainer*/, name/*:string*/, content/*:?RawBytes*/, opts/*:?any*/)/*:CFBEntry*/ {
-	init_cfb(cfb);
-	var file = CFB.find(cfb, name);
+	var unsafe = opts && opts.unsafe;
+	if(!unsafe) init_cfb(cfb);
+	var file = !unsafe && CFB.find(cfb, name);
 	if(!file) {
 		var fpath/*:string*/ = cfb.FullPaths[0];
 		if(name.slice(0, fpath.length) == fpath) fpath = name;
@@ -696,7 +697,7 @@ function cfb_add(cfb/*:CFBContainer*/, name/*:string*/, content/*:?RawBytes*/, o
 		file = ({name: filename(name), type: 2}/*:any*/);
 		cfb.FileIndex.push(file);
 		cfb.FullPaths.push(fpath);
-		CFB.utils.cfb_gc(cfb);
+		if(!unsafe) CFB.utils.cfb_gc(cfb);
 	}
 	/*:: if(!file) throw new Error("unreachable"); */
 	file.content = (content/*:any*/);

package.json

@@ -1,6 +1,6 @@
 {
   "name": "cfb",
-  "version": "1.0.5",
+  "version": "1.0.6",
   "author": "sheetjs",
   "description": "Compound File Binary File Format extractor",
   "keywords": [

shim.js

@@ -18,6 +18,12 @@ if(!Array.prototype.indexOf) Array.prototype.indexOf = function(needle) {
   return -1;
 };
 
+if(!Array.prototype.lastIndexOf) Array.prototype.lastIndexOf = function(needle) {
+  var len = (this.length>>>0), i = len - 1;
+  for(; i>=0; --i) if(this[i] === needle) return i;
+  return -1;
+};
+
 if(!Array.isArray) Array.isArray = function(obj) { return Object.prototype.toString.call(obj) === "[object Array]"; };
 
 if(typeof ArrayBuffer !== 'undefined' && !ArrayBuffer.prototype.slice) ArrayBuffer.prototype.slice = function(start, end) {

types/index.d.ts

@@ -91,10 +91,16 @@ export interface CFB$Container {
     };
 }
 
+/** cfb_add options */
+export interface CFB$AddOpts {
+    /** Skip existence and safety checks (best for bulk write operations) */
+    unsafe?: boolean;
+}
+
 /** General utilities */
 export interface CFB$Utils {
     cfb_new(opts?: any): CFB$Container;
-    cfb_add(cfb: CFB$Container, name: string, content: any, opts?: any): CFB$Entry;
+    cfb_add(cfb: CFB$Container, name: string, content: any, opts?: CFB$AddOpts): CFB$Entry;
     cfb_del(cfb: CFB$Container, name: string): boolean;
     cfb_mov(cfb: CFB$Container, old_name: string, new_name: string): boolean;
     cfb_gc(cfb: CFB$Container): void;

types/roundtrip.ts

@@ -61,7 +61,7 @@ const old_cfb = CFB.read("t2.xls", {type:"file"});
 const new_cfb = CFB.utils.cfb_new({root:"R", clsid: old_cfb.FileIndex[0].clsid });
 old_cfb.FullPaths.forEach((p, i) => {
 	if(p.slice(-1) === "/") return;
-	CFB.utils.cfb_add(new_cfb, p.replace(/^[^/]*/,"R"), old_cfb.FileIndex[i].content);
+	CFB.utils.cfb_add(new_cfb, p.replace(/^[^/]*/,"R"), old_cfb.FileIndex[i].content, {unsafe: true});
 });
 dumpit(new_cfb);
 CFB.writeFile(new_cfb, "t3.xls");

xlscfb.flow.js

@@ -38,7 +38,7 @@ type CFBFiles = {[n:string]:CFBEntry};
 /* [MS-CFB] v20171201 */
 var CFB = (function _CFB(){
 var exports/*:CFBModule*/ = /*::(*/{}/*:: :any)*/;
-exports.version = '1.0.5';
+exports.version = '1.0.6';
 /* [MS-CFB] 2.6.4 */
 function namecmp(l/*:string*/, r/*:string*/)/*:number*/ {
 	var L = l.split("/"), R = r.split("/");
@@ -216,8 +216,8 @@ function build_full_paths(FI/*:CFBFileIndex*/, FP/*:Array<string>*/, Paths/*:Arr
 			if(R !== -1 && dad[R] !== R) dad[i] = dad[R];
 		}
 		if(C !== -1 /*NOSTREAM*/) dad[C] = i;
-		if(L !== -1) { dad[L] = dad[i]; q.push(L); }
-		if(R !== -1) { dad[R] = dad[i]; q.push(R); }
+		if(L !== -1) { dad[L] = dad[i]; if(q.lastIndexOf(L) < j) q.push(L); }
+		if(R !== -1) { dad[R] = dad[i]; if(q.lastIndexOf(R) < j) q.push(R); }
 	}
 	for(i=1; i < pl; ++i) if(dad[i] === i) {
 		if(R !== -1 /*NOSTREAM*/ && dad[R] !== R) dad[i] = dad[R];
@@ -684,8 +684,9 @@ function cfb_new(opts/*:?any*/)/*:CFBContainer*/ {
 }
 
 function cfb_add(cfb/*:CFBContainer*/, name/*:string*/, content/*:?RawBytes*/, opts/*:?any*/)/*:CFBEntry*/ {
-	init_cfb(cfb);
-	var file = CFB.find(cfb, name);
+	var unsafe = opts && opts.unsafe;
+	if(!unsafe) init_cfb(cfb);
+	var file = !unsafe && CFB.find(cfb, name);
 	if(!file) {
 		var fpath/*:string*/ = cfb.FullPaths[0];
 		if(name.slice(0, fpath.length) == fpath) fpath = name;
@@ -696,7 +697,7 @@ function cfb_add(cfb/*:CFBContainer*/, name/*:string*/, content/*:?RawBytes*/, o
 		file = ({name: filename(name), type: 2}/*:any*/);
 		cfb.FileIndex.push(file);
 		cfb.FullPaths.push(fpath);
-		CFB.utils.cfb_gc(cfb);
+		if(!unsafe) CFB.utils.cfb_gc(cfb);
 	}
 	/*:: if(!file) throw new Error("unreachable"); */
 	file.content = (content/*:any*/);

xlscfb.js

@@ -8,7 +8,7 @@ var DO_NOT_EXPORT_CFB = true;
 /* [MS-CFB] v20171201 */
 var CFB = (function _CFB(){
 var exports = {};
-exports.version = '1.0.5';
+exports.version = '1.0.6';
 /* [MS-CFB] 2.6.4 */
 function namecmp(l, r) {
 	var L = l.split("/"), R = r.split("/");
@@ -186,8 +186,8 @@ function build_full_paths(FI, FP, Paths) {
 			if(R !== -1 && dad[R] !== R) dad[i] = dad[R];
 		}
 		if(C !== -1 /*NOSTREAM*/) dad[C] = i;
-		if(L !== -1) { dad[L] = dad[i]; q.push(L); }
-		if(R !== -1) { dad[R] = dad[i]; q.push(R); }
+		if(L !== -1) { dad[L] = dad[i]; if(q.lastIndexOf(L) < j) q.push(L); }
+		if(R !== -1) { dad[R] = dad[i]; if(q.lastIndexOf(R) < j) q.push(R); }
 	}
 	for(i=1; i < pl; ++i) if(dad[i] === i) {
 		if(R !== -1 /*NOSTREAM*/ && dad[R] !== R) dad[i] = dad[R];
@@ -648,8 +648,9 @@ function cfb_new(opts) {
 }
 
 function cfb_add(cfb, name, content, opts) {
-	init_cfb(cfb);
-	var file = CFB.find(cfb, name);
+	var unsafe = opts && opts.unsafe;
+	if(!unsafe) init_cfb(cfb);
+	var file = !unsafe && CFB.find(cfb, name);
 	if(!file) {
 		var fpath = cfb.FullPaths[0];
 		if(name.slice(0, fpath.length) == fpath) fpath = name;
@@ -660,7 +661,7 @@ function cfb_add(cfb, name, content, opts) {
 		file = ({name: filename(name), type: 2});
 		cfb.FileIndex.push(file);
 		cfb.FullPaths.push(fpath);
-		CFB.utils.cfb_gc(cfb);
+		if(!unsafe) CFB.utils.cfb_gc(cfb);
 	}
 file.content = (content);
 	file.size = content ? content.length : 0;