Ticket #145: chm.diff

chm.py

@@ -6 +6 @@
-  http://www.wotsit.org (search "chm")
-- chmlib library
-  http://www.jedrea.com/chmlib/
+- Unofficial CHM Spec
+  http://savannah.nongnu.org/projects/chmspec
+- Microsoft's HTML Help (.chm) format
+  http://www.speakeasy.org/~russotto/chm/chmformat.html
-
-Author: Victor Stinner
-Creation date: 2007-03-04
@@ -13 +17 @@
-
-from hachoir_parser import Parser
-from hachoir_core.field import (Field, FieldSet, ParserError,
-
+16, UInt
-    RawBytes, PaddingBytes,
-    Enum, String)
-from hachoir_core.endian import LITTLE_ENDIAN
@@ -42 +46 @@
-                raise ParserError("CHM: CWord is limited to 64 bits")
-            addr += 8
-            byte = stream.readBits(addr, 8, endian)
+        value <<= 7
-        value += byte
-        self.createValue = lambda: value
-
@@ -99 +104 @@
-    def createFields(self):
-        yield CWord(self, "name_len")
-        yield String(self, "name", self["name_len"].value, charset="UTF-8")
-pace
-
-
+ection", "Section number that the entry data is in.
+, "Start offset of the data"
+, "Length of the data"
-
-    def createDescription(self):
-        return "%s (%s)" % (self["name"].value, self["length"].display)
@@ -118 +123 @@
-
-        # Entries
-        stop = self.size - self["free_space"].value * 8
+        entry_count = 0
-        while self.current_size < stop:
-            yield PMGL_Entry(self, "entry[]")
+            entry_count+=1
-
-        # Padding
-
+
+
+
+
-        if padding:
-            yield PaddingBytes(self, "padding", padding)
+        for i in range(num_quickref*quickref_frequency, 0, -quickref_frequency):
+            yield UInt16(self, "quickref[%i]"%i)
+        yield UInt16(self, "entry_count")
-
-class PMGI_Entry(FieldSet):
-    def createFields(self):
@@ -158 +171 @@
-
-        if nb_dir < 0:
-            nb_dir = 1
+        entries=[]
-        for index in xrange(nb_dir):
-
+
+
+
-
+        self.entries=entries
-        if self.current_size < self.size:
-            yield PMGI(self, "pmgi", size=block_size)
-
+class NameList(FieldSet):
+    def createFields(self):
+        yield UInt16(self, "length", "Length of name list in 2-byte blocks")
+        yield UInt16(self, "count", "Number of entries in name list")
+        for index in range(self["count"].value):
+            length=UInt16(self, "name_len[]", "Length of name in 2-byte blocks, excluding terminating null")
+            yield length
+            yield String(self, "name[]", length.value*2+2, charset="UTF-16-LE")
+
+class ControlData(FieldSet):
+    def createFields(self):
+        yield UInt32(self, "count", "Number of DWORDS in this struct")
+        yield String(self, "type", 4, "Type of compression")
+        if self["type"].value!='LZXC': return
+        yield UInt32(self, "version", "Compression version")
+        version=self["version"].value
+        if version==1: block='bytes'
+        else: block='32KB blocks'
+        yield UInt32(self, "reset_interval", "LZX: Reset interval in %s"%block)
+        yield UInt32(self, "window_size", "LZX: Window size in %s"%block)
+        yield UInt32(self, "cache_size", "LZX: Cache size in %s"%block)
+        yield UInt32(self, "unknown[]")
+
+class ResetTable(FieldSet):
+    def createFields(self):
+        yield UInt32(self, "unknown[]", "Version number?")
+        yield UInt32(self, "count", "Number of entries")
+        yield UInt32(self, "entry_size", "Size of each entry")
+        yield UInt32(self, "header_size", "Size of this header")
+        yield UInt64(self, "uncompressed_size")
+        yield UInt64(self, "compressed_size")
+        yield UInt64(self, "block_size", "Block size in bytes")
+        for i in xrange(self["count"].value):
+            yield UInt64(self, "block_location[]", "location in compressed data of 1st block boundary in uncompressed data")
+
+class SystemEntry(FieldSet):
+    ENTRY_TYPE={0:"HHP: [OPTIONS]: Contents File",
+                1:"HHP: [OPTIONS]: Index File",
+                2:"HHP: [OPTIONS]: Default Topic",
+                3:"HHP: [OPTIONS]: Title",
+                4:"File Metadata",
+                5:"HHP: [OPTIONS]: Default Window",
+                6:"HHP: [OPTIONS]: Compiled file",
+                # 7 present only in files with Binary Index; unknown function
+                # 8 unknown function
+                9: "Version",
+                10: "Timestamp",
+                # 11 only in Binary TOC files
+                12: "Number of Info Types",
+                13: "#IDXHDR file",
+                # 14 unknown function
+                # 15 checksum??
+                16:"HHP: [OPTIONS]: Default Font",
+    }
+    def createFields(self):
+        yield Enum(UInt16(self, "type", "Type of entry"),self.ENTRY_TYPE)
+        yield UInt16(self, "length", "Length of entry")
+        yield RawBytes(self, "data", self["length"].value)
+    def createDescription(self):
+        return '#SYSTEM Entry, Type %s'%self["type"].display
+        
+class SystemFile(FieldSet):
+    def createFields(self):
+        yield UInt32(self, "version", "Either 2 or 3")
+        while self.current_size < self.size:
+            yield SystemEntry(self, "entry[]")
+
-class ChmFile(Parser):
-    tags = {
-        "id": "chm",
@@ -176 +260 @@
-    endian = LITTLE_ENDIAN
-
-    def validate(self):
-.stream.readBytes(0, 4)
+["itsf/magic"].value
-            return "Invalid magic"
-        if self["itsf/version"].value != 3:
-            return "Invalid version"
@@ -189 +273 @@
-        padding = self.seekByte(self["itsf/dir_offset"].value)
-        if padding:
-            yield padding
-
+
+
-
+        # alternately, this could be a SeekableFieldSet, which would eliminate
+        # the need for running through all entries and would produce them in
+        # the order that they appear in the index. However, then, it would not
+        # be possible to see every byte of the file, as some may be skipped over.
+        for field in directory:
+            pass # exhaust directory to read all entries in order to sort them
+
+        entries=sorted(directory.entries,key=lambda x:x["start"].value)
+        otherentries={}
+        for entry in entries:
+            if entry["section"].value != 0:
+                otherentries.setdefault(entry["section"].value,[]).append(entry)
+                continue
+            if entry["length"].value==0: continue
+            padding=self.seekByte(self["itsf/data_offset"].value+entry["start"].value)
+            if padding:
+                yield padding
+            name=entry["name"].value
+            if name=="::DataSpace/NameList":
+                yield NameList(self, "name_list")
+            elif name.startswith('::DataSpace/Storage/'):
+                sectname=str(name.split('/')[2])
+                if name.endswith('/SpanInfo'):
+                    yield UInt64(self, "%s_spaninfo"%sectname, "Size of uncompressed data in the %s section"%sectname)
+                elif name.endswith('/ControlData'):
+                    yield ControlData(self, "%s_controldata"%sectname, "Data about the compression scheme", size=entry["length"].value*8)
+                elif name.endswith('/Transform/List'):
+                    yield String(self, "%s_transform_list"%sectname, 38, description="Transform/List element", charset="UTF-16-LE")
+                elif name.endswith('/Transform/{7FC28940-9D31-11D0-9B27-00A0C91E9C7C}/InstanceData/ResetTable'):
+                    yield ResetTable(self, "%s_reset_table"%sectname, "LZX Reset Table", size=entry["length"].value*8)
+                elif name.endswith('/Content'):
+                    # eventually, a LZX wrapper will appear here, we hope!
+                    yield RawBytes(self, "%s_content"%sectname, entry["length"].value, "Content for the %s section"%sectname)
+                else:
+                    yield RawBytes(self, "entry_data[]", entry["length"].value, name)
+            elif name=="/#SYSTEM":
+                yield SystemFile(self, "system_file", size=entry["length"].value*8)
+            else:
+                yield RawBytes(self, "entry_data[]", entry["length"].value, name)
+
-        size = (self.size - self.current_size) // 8
-        if size:
-            yield RawBytes(self, "raw_end", size)
-
+    def getFile(self, filename):
+        page=0
+        if 'pmgi' in self['/dir']:
+            for entry in self['/dir/pmgi'].array('entry'):
+                if entry['name'].value <= filename:
+                    page=entry['page'].value
+        pmgl=self['/dir/pmgl[%i]'%page]
+        for entry in pmgl.array('entry'):
+            if entry['name'].value == filename:
+                return entry
+        raise ParserError("File '%s' not found!"%filename)
+
-    def createContentSize(self):
-        return self["file_size/file_size"].value * 8
-