fix(VR): added support for specific character set (#291)

* added support for specific character set refactored VR classes refactored reading/writing ascii and encoded strings changed hex writing to direct uint8 writing reimplemented null padded read removed unused code unified test data loading and cache the test data locally for faster test runs * re-enabled test and removed wrong only * fixed test case that was overlooked Co-authored-by: flink <[email protected]>
dcmjs-org · Jul 10, 2022 · bfd6693 · bfd6693
1 parent a0a0fd5
commit bfd6693
Show file tree

Hide file tree

Showing 9 changed files with 390 additions and 344 deletions.
diff --git a/package.json b/package.json
@@ -8,7 +8,7 @@
     "example": "examples"
   },
   "scripts": {
-    "test": "jest",
+    "test": "jest --testTimeout 60000",
     "build": "rollup -c",
     "build:examples": "npm run build && npx cpx 'build/**/*.{js,map}' examples/js",
     "start": "rollup -c -w",

diff --git a/src/BufferStream.js b/src/BufferStream.js
@@ -1,38 +1,3 @@
-//http://jonisalonen.com/2012/from-utf-16-to-utf-8-in-javascript/
-function toUTF8Array(str) {
-    var utf8 = [];
-    for (var i = 0; i < str.length; i++) {
-        var charcode = str.charCodeAt(i);
-        if (charcode < 0x80) utf8.push(charcode);
-        else if (charcode < 0x800) {
-            utf8.push(0xc0 | (charcode >> 6), 0x80 | (charcode & 0x3f));
-        } else if (charcode < 0xd800 || charcode >= 0xe000) {
-            utf8.push(
-                0xe0 | (charcode >> 12),
-                0x80 | ((charcode >> 6) & 0x3f),
-                0x80 | (charcode & 0x3f)
-            );
-        }
-        // surrogate pair
-        else {
-            i++;
-            // UTF-16 encodes 0x10000-0x10FFFF by
-            // subtracting 0x10000 and splitting the
-            // 20 bits of 0x0-0xFFFFF into two halves
-            charcode =
-                0x10000 +
-                (((charcode & 0x3ff) << 10) | (str.charCodeAt(i) & 0x3ff));
-            utf8.push(
-                0xf0 | (charcode >> 18),
-                0x80 | ((charcode >> 12) & 0x3f),
-                0x80 | ((charcode >> 6) & 0x3f),
-                0x80 | (charcode & 0x3f)
-            );
-        }
-    }
-    return utf8;
-}
-
 function toInt(val) {
     if (isNaN(val)) {
         throw new Error("Not a number: " + val);
@@ -62,6 +27,7 @@ class BufferStream {
         this.offset = 0;
         this.isLittleEndian = littleEndian || false;
         this.size = 0;
+        this.encoder = new TextEncoder("utf-8");
     }
 
     setEndian(isLittle) {
@@ -74,6 +40,15 @@ class BufferStream {
         return this.increment(1);
     }
 
+    writeUint8Repeat(value, count) {
+        const v = toInt(value);
+        this.checkSize(count);
+        for (let i = 0; i < count; i++) {
+            this.view.setUint8(this.offset + i, v);
+        }
+        return this.increment(count);
+    }
+
     writeInt8(value) {
         this.checkSize(1);
         this.view.setInt8(this.offset, toInt(value));
@@ -129,40 +104,23 @@ class BufferStream {
         return this.increment(8);
     }
 
-    writeString(value) {
-        value = value || "";
-        var utf8 = toUTF8Array(value),
-            bytelen = utf8.length;
+    writeUTF8String(value) {
+        const encodedString = this.encoder.encode(value);
+        this.checkSize(encodedString.byteLength);
+        new Uint8Array(this.buffer).set(encodedString, this.offset);
+        return this.increment(encodedString.byteLength);
+    }
 
-        this.checkSize(bytelen);
+    writeAsciiString(value) {
+        value = value || "";
+        var len = value.length;
+        this.checkSize(len);
         var startOffset = this.offset;
-        for (var i = 0; i < bytelen; i++) {
-            this.view.setUint8(startOffset, utf8[i]);
-            startOffset++;
+        for (let i = 0; i < len; i++) {
+            var charcode = value.charCodeAt(i);
+            this.view.setUint8(startOffset + i, charcode);
         }
-        return this.increment(bytelen);
-    }
-
-    writeHex(value) {
-        var len = value.length,
-            blen = len / 2,
-            startOffset = this.offset;
-        this.checkSize(blen);
-        for (var i = 0; i < len; i += 2) {
-            var code = parseInt(value[i], 16),
-                nextCode;
-            if (i == len - 1) {
-                nextCode = null;
-            } else {
-                nextCode = parseInt(value[i + 1], 16);
-            }
-            if (nextCode !== null) {
-                code = (code << 4) | nextCode;
-            }
-            this.view.setUint8(startOffset, code);
-            startOffset++;
-        }
-        return this.increment(blen);
+        return this.increment(len);
     }
 
     readUint32() {
@@ -183,6 +141,10 @@ class BufferStream {
         return val;
     }
 
+    peekUint8(offset) {
+        return this.view.getUint8(this.offset + offset);
+    }
+
     readUint8Array(length) {
         var arr = new Uint8Array(this.buffer, this.offset, length);
         this.increment(length);
@@ -224,18 +186,36 @@ class BufferStream {
         return val;
     }
 
-    readString(length) {
-        var chars = [];
+    readAsciiString(length) {
+        var result = "";
         var start = this.offset;
         var end = this.offset + length;
         if (end >= this.buffer.byteLength) {
             end = this.buffer.byteLength;
         }
         for (let i = start; i < end; ++i) {
-            chars.push(String.fromCharCode(this.view.getUint8(i)));
-            this.increment(1);
+            result += String.fromCharCode(this.view.getUint8(i));
+        }
+        this.increment(end - start);
+        return result;
+    }
+
+    readVR() {
+        var vr =
+            String.fromCharCode(this.view.getUint8(this.offset)) +
+            String.fromCharCode(this.view.getUint8(this.offset + 1));
+        this.increment(2);
+        return vr;
+    }
+
+    readEncodedString(length) {
+        if (this.offset + length >= this.buffer.byteLength) {
+            length = this.buffer.byteLength - this.offset;
         }
-        return chars.join("");
+        const view = new DataView(this.buffer, this.offset, length);
+        const result = this.decoder.decode(view);
+        this.increment(length);
+        return result;
     }
 
     readHex(length) {
@@ -354,6 +334,11 @@ class ReadBufferStream extends BufferStream {
         this.noCopy = options.noCopy;
         this.startOffset = this.offset;
         this.endOffset = this.size;
+        this.decoder = new TextDecoder("latin1");
+    }
+
+    setDecoder(decoder) {
+        this.decoder = decoder;
     }
 
     getBuffer(start, end) {
@@ -368,20 +353,6 @@ class ReadBufferStream extends BufferStream {
         return this.buffer.slice(start, end);
     }
 
-    readString(length) {
-        var chars = [];
-        var start = this.offset;
-        var end = this.offset + length;
-        if (end >= this.endOffset) {
-            end = this.endOffset;
-        }
-        for (let i = start; i < end; ++i) {
-            chars.push(String.fromCharCode(this.view.getUint8(i)));
-            this.increment(1);
-        }
-        return chars.join("");
-    }
-
     reset() {
         this.offset = this.startOffset;
         return this;
@@ -399,6 +370,10 @@ class ReadBufferStream extends BufferStream {
         throw new Error(value, "writeUint8 not implemented");
     }
 
+    writeUint8Repeat(value, count) {
+        throw new Error(value, "writeUint8Repeat not implemented");
+    }
+
     writeInt8(value) {
         throw new Error(value, "writeInt8 not implemented");
     }
@@ -431,12 +406,12 @@ class ReadBufferStream extends BufferStream {
         throw new Error(value, "writeDouble not implemented");
     }
 
-    writeString(value) {
-        throw new Error(value, "writeString not implemented");
+    writeAsciiString(value) {
+        throw new Error(value, "writeAsciiString not implemented");
     }
 
-    writeHex(value) {
-        throw new Error(value, "writeHex not implemented");
+    writeUTF8String(value) {
+        throw new Error(value, "writeUTF8String not implemented");
     }
 
     checkSize(step) {

diff --git a/src/DicomDict.js b/src/DicomDict.js
@@ -20,8 +20,8 @@ class DicomDict {
     write(writeOptions = { allowInvalidVRLength: false }) {
         var metaSyntax = EXPLICIT_LITTLE_ENDIAN;
         var fileStream = new WriteBufferStream(4096, true);
-        fileStream.writeHex("00".repeat(128));
-        fileStream.writeString("DICM");
+        fileStream.writeUint8Repeat(0, 128);
+        fileStream.writeAsciiString("DICM");
 
         var metaStream = new WriteBufferStream(1024);
         if (!this.meta["00020010"]) {

diff --git a/src/DicomMessage.js b/src/DicomMessage.js
@@ -9,6 +9,11 @@ const EXPLICIT_LITTLE_ENDIAN = "1.2.840.10008.1.2.1";
 const EXPLICIT_BIG_ENDIAN = "1.2.840.10008.1.2.2";
 const singleVRs = ["SQ", "OF", "OW", "OB", "UN", "LT"];
 
+const encodingMapping = {
+    "iso-ir-192": "utf-8",
+    "": "latin1"
+};
+
 const encapsulatedSyntaxes = [
     "1.2.840.10008.1.2.4.50",
     "1.2.840.10008.1.2.4.51",
@@ -79,7 +84,30 @@ class DicomMessage {
                     options
                 );
                 const cleanTagString = readInfo.tag.toCleanString();
-
+                if (cleanTagString === "00080005") {
+                    if (readInfo.values.length > 0) {
+                        let coding = readInfo.values[0];
+                        coding = coding
+                            .replaceAll("_", "-")
+                            .replaceAll(" ", "-")
+                            .toLowerCase();
+                        if (coding in encodingMapping) {
+                            coding = encodingMapping[coding];
+                        }
+                        try {
+                            bufferStream.setDecoder(new TextDecoder(coding));
+                        } catch (error) {
+                            console.warn(error);
+                        }
+                    }
+                    if (readInfo.values.length > 1) {
+                        console.warn(
+                            "multiple encodings not supported, using first encoding!",
+                            readInfo.values
+                        );
+                    }
+                    readInfo.values = ["ISO_IR 192"]; // change SpecificCharacterSet to UTF-8
+                }
                 dict[cleanTagString] = {
                     vr: readInfo.vr.type,
                     Value: readInfo.values
@@ -130,7 +158,7 @@ class DicomMessage {
             useSyntax = EXPLICIT_LITTLE_ENDIAN;
         stream.reset();
         stream.increment(128);
-        if (stream.readString(4) !== "DICM") {
+        if (stream.readAsciiString(4) !== "DICM") {
             throw new Error("Invalid a dicom file");
         }
         var el = DicomMessage._readTag(stream, useSyntax),
@@ -161,7 +189,7 @@ class DicomMessage {
         var written = 0;
 
         var sortedTags = Object.keys(jsonObjects).sort();
-        sortedTags.forEach(function(tagString) {
+        sortedTags.forEach(function (tagString) {
             var tag = Tag.fromString(tagString),
                 tagObject = jsonObjects[tagString],
                 vrType = tagObject.vr,
@@ -230,7 +258,7 @@ class DicomMessage {
             }
             vr = ValueRepresentation.createByTypeString(vrType);
         } else {
-            vrType = stream.readString(2);
+            vrType = stream.readVR();
             vr = ValueRepresentation.createByTypeString(vrType);
             if (vr.isExplicit()) {
                 stream.increment(2);

diff --git a/src/Tag.js b/src/Tag.js
@@ -17,37 +17,17 @@ class Tag {
     toString() {
         return (
             "(" +
-            paddingLeft(
-                "0000",
-                this.group()
-                    .toString(16)
-                    .toUpperCase()
-            ) +
+            paddingLeft("0000", this.group().toString(16).toUpperCase()) +
             "," +
-            paddingLeft(
-                "0000",
-                this.element()
-                    .toString(16)
-                    .toUpperCase()
-            ) +
+            paddingLeft("0000", this.element().toString(16).toUpperCase()) +
             ")"
         );
     }
 
     toCleanString() {
         return (
-            paddingLeft(
-                "0000",
-                this.group()
-                    .toString(16)
-                    .toUpperCase()
-            ) +
-            paddingLeft(
-                "0000",
-                this.element()
-                    .toString(16)
-                    .toUpperCase()
-            )
+            paddingLeft("0000", this.group().toString(16).toUpperCase()) +
+            paddingLeft("0000", this.element().toString(16).toUpperCase())
         );
     }
 
@@ -147,12 +127,12 @@ class Tag {
             written += 4;
         } else {
             if (vr.isExplicit()) {
-                stream.writeString(vr.type);
-                stream.writeHex("0000");
+                stream.writeAsciiString(vr.type);
+                stream.writeUint16(0);
                 stream.writeUint32(valueLength);
                 written += 8;
             } else {
-                stream.writeString(vr.type);
+                stream.writeAsciiString(vr.type);
                 stream.writeUint16(valueLength);
                 written += 4;
             }