[pywin32-checkins] pywin32/com/win32comext/mapi/src mapi.i,1.2,1.3 mapilib.i,1.1,1.2
OLD project page for the Python extensions for Windows
Brought to you by:
mhammond
Menu
▾
▴
[pywin32-checkins] pywin32/com/win32comext/mapi/src mapi.i,1.2,1.3 mapilib.i,1.1,1.2
|
From: <mha...@us...> - 2003-09-01 08:05:51
|
Update of /cvsroot/pywin32/pywin32/com/win32comext/mapi/src
In directory sc8-pr-cvs1:/tmp/cvs-serv27409
Modified Files:
mapi.i mapilib.i
Log Message:
Add WrapCompressedRTFStream, and a little utility for converting
RTF -> HTML (but only when the source of the RTF was originally HTML -
which begs the question why MS didn't give access to the HTML!)
Index: mapi.i
===================================================================
RCS file: /cvsroot/pywin32/pywin32/com/win32comext/mapi/src/mapi.i,v
retrieving revision 1.2
retrieving revision 1.3
diff -C2 -d -r1.2 -r1.3
*** mapi.i 25 Jan 2000 09:59:57 -0000 1.2
--- mapi.i 1 Sep 2003 08:05:36 -0000 1.3
***************
*** 482,485 ****
--- 482,492 ----
);
+ // @pyswig <PyIStream>|WrapCompressedRTFStream|
+ HRESULT WrapCompressedRTFStream(
+ IStream *INPUT, // @pyparm <o PyIStream>|stream||Message stream
+ unsigned long ulflags, // @pyparm int|flags||
+ IStream **OUTPUT
+ );
+
%native(MAPIUIDFromBinary) MAPIUIDFromBinary;
%{
***************
*** 596,597 ****
--- 603,758 ----
}
%}
+
+ %{
+ // Code for converting RTF to HTML.
+ // Found at http://www.wischik.com/lu/programmer/mapi_utils.html
+ // MarkH converted it to Python, but was too slow. Moving to a regex
+ // based parser was too much work.
+
+ // DECODERTFHTML -- Given an uncompressed RTF body of the message,
+ // and assuming that it contains encoded-html, this function
+ // turns it onto regular html.
+ // [in] (buf,*len) indicate the start and length of the uncompressed RTF body.
+ // [out] the buffer is overwritten with the HTML version, null-terminated,
+ // and *len indicates the length of this HTML.
+ //
+ // Notes: (1) because of how the encoding works, the HTML version is necessarily
+ // shorter than the encoded version. That's why it's safe for the function to
+ // place the decoded html in the same buffer that formerly held the encoded stuff.
+ // (2) Some messages include characters \'XX, where XX is a hexedecimal number.
+ // This function simply converts this into ASCII. The conversion will only make
+ // sense if the right code-page is being used. I don't know how rtf specifies which
+ // code page it wants.
+ // (3) By experiment, I discovered that \pntext{..} and \liN and \fi-N are RTF
+ // markup that should be removed. There might be other RTF markup that should
+ // also be removed. But I don't know what else.
+ //
+ void decodertfhtml(char *buf,unsigned int *len)
+ { // c -- pointer to where we're reading from
+ // d -- pointer to where we're writing to. Invariant: d<c
+ // max -- how far we can read from (i.e. to the end of the original rtf)
+ // ignore_tag -- stores 'N': after \mhtmlN, we will ignore the subsequent \htmlN.
+ char *c=buf, *max=buf+*len, *d=buf; int ignore_tag=-1;
+ // First, we skip forwards to the first \htmltag.
+ while (c<max && strncmp(c,"{\\*\\htmltag",11)!=0) c++;
+ //
+ // Now work through the document. Our plan is as follows:
+ // * Ignore { and }. These are part of RTF markup.
+ // * Ignore \htmlrtf...\htmlrtf0. This is how RTF keeps its equivalent markup separate from the html.
+ // * Ignore \r and \n. The real carriage returns are stored in \par tags.
+ // * Ignore \pntext{..} and \liN and \fi-N. These are RTF junk.
+ // * Convert \par and \tab into \r\n and \t
+ // * Convert \'XX into the ascii character indicated by the hex number XX
+ // * Convert \{ and \} into { and }. This is how RTF escapes its curly braces.
+ // * When we get \*\mhtmltagN, keep the tag, but ignore the subsequent \*\htmltagN
+ // * When we get \*\htmltagN, keep the tag as long as it isn't subsequent to a \*\mhtmltagN
+ // * All other text should be kept as it is.
+ while (c<max)
+ { if (*c=='{') c++;
+ else if (*c=='}') c++;
+ else if (strncmp(c,"\\*\\htmltag",10)==0)
+ { c+=10; int tag=0; while (*c>='0' && *c<='9') {tag=tag*10+*c-'0'; c++;}
+ if (*c==' ') c++;
+ if (tag==ignore_tag) {while (c<max && *c!='}') c++; if (*c=='}') c++;}
+ ignore_tag=-1;
+ }
+ else if (strncmp(c,"\\*\\mhtmltag",11)==0)
+ { c+=11; int tag=0; while (*c>='0' && *c<='9') {tag=tag*10+*c-'0'; c++;}
+ if (*c==' ') c++;
+ ignore_tag=tag;
+ }
+ else if (strncmp(c,"\\par",4)==0) {strcpy(d,"\r\n"); d+=2; c+=4; if (*c==' ') c++;}
+ else if (strncmp(c,"\\tab",4)==0) {strcpy(d," "); d+=3; c+=4; if (*c==' ') c++;}
+ else if (strncmp(c,"\\li",3)==0)
+ { c+=3; while (*c>='0' && *c<='9') c++; if (*c==' ') c++;
+ }
+ else if (strncmp(c,"\\fi-",4)==0)
+ { c+=4; while (*c>='0' && *c<='9') c++; if (*c==' ') c++;
+ }
+ else if (strncmp(c,"\\'",2)==0)
+ { unsigned int hi=c[2], lo=c[3];
+ if (hi>='0' && hi<='9') hi-='0'; else if (hi>='A' && hi<='Z') hi=hi-'A'+10; else if (hi>='a' && hi<='z') hi=hi-'a'+10;
+ if (lo>='0' && lo<='9') lo-='0'; else if (lo>='A' && lo<='Z') lo=lo-'A'+10; else if (lo>='a' && lo<='z') lo=lo-'a'+10;
+ *((unsigned char*)d) = (unsigned char)(hi*16+lo);
+ c+=4; d++;
+ }
+ else if (strncmp(c,"\\pntext",7)==0) {c+=7; while (c<max && *c!='}') c++;}
+ else if (strncmp(c,"\\htmlrtf",8)==0)
+ { c++; while (c<max && strncmp(c,"\\htmlrtf0",9)!=0) c++;
+ if (c<max) c+=9; if (*c==' ') c++;
+ }
+ else if (*c=='\r' || *c=='\n') c++;
+ else if (strncmp(c,"\\{",2)==0) {*d='{'; d++; c+=2;}
+ else if (strncmp(c,"\\}",2)==0) {*d='}'; d++; c+=2;}
+ else {*d=*c; c++; d++;}
+ }
+ *d=0; d++;
+ *len = (unsigned int)(d-buf);
+ }
+
+
+ bool isrtfhtml(const char *buf,unsigned int len)
+ { // We look for the words "\fromhtml" somewhere in the file.
+ // If the rtf encodes text rather than html, then instead
+ // it will only find "\fromtext".
+ for (const char *c=buf; c<buf+len; c++)
+ { if (strncmp(c,"\\from",5)==0) return strncmp(c,"\\fromhtml",9)==0;
+ } return false;
+ }
+
+ // @pyswig |RTFStreamToHTML|
+ static PyObject *MyRTFStreamToHTML(PyObject *self, PyObject *args)
+ {
+ PyObject *obStream;
+ HRESULT hr;
+ // @pyparm <o PyIStream>|The stream to read the uncompressed RTF from||
+ if (!PyArg_ParseTuple(args, "O:RTFStreamToHTML", &obStream))
+ return NULL;
+ IStream *pStream = NULL;
+
+ if (!PyCom_InterfaceFromPyObject(obStream, IID_IStream, (void **)&pStream, FALSE))
+ return NULL;
+
+ // all exit from here via 'exit', and no Python until POSTCALL
+ PY_INTERFACE_PRECALL;
+ PyObject *ret = NULL;
+ unsigned int bufsize=10240;
+ char *htmlbuf = (char *)malloc(bufsize);
+ unsigned int htmlsize=0; bool done=(htmlbuf==NULL);
+ while (!done)
+ { ULONG red; hr = pStream->Read(htmlbuf+htmlsize, bufsize-htmlsize, &red);
+ if (hr!=S_OK) {htmlbuf[htmlsize]=0; done=true;}
+ else
+ { htmlsize+=red; done = (red < bufsize-htmlsize);
+ if (!done)
+ { unsigned int newsize=2*htmlsize;
+ htmlbuf = (char *)realloc(htmlbuf, newsize);
+ bufsize=newsize;
+ }
+ }
+ }
+ bool ok;
+ if (htmlbuf) {
+ ok = isrtfhtml(htmlbuf,htmlsize);
+ if (ok)
+ decodertfhtml(htmlbuf,&htmlsize);
+ }
+ PY_INTERFACE_POSTCALL;
+ if (htmlbuf==0) {
+ PyErr_NoMemory();
+ goto exit;
+ }
+ if (!ok) {
+ Py_INCREF(Py_None);
+ ret = Py_None;
+ goto exit;
+ }
+ ret = PyString_FromStringAndSize(htmlbuf, htmlsize-1);
+ exit:
+ if (pStream) pStream->Release();
+ if (htmlbuf)
+ free(htmlbuf);
+ return ret;
+ }
+ %}
+ %native(RTFStreamToHTML) MyRTFStreamToHTML;
Index: mapilib.i
===================================================================
RCS file: /cvsroot/pywin32/pywin32/com/win32comext/mapi/src/mapilib.i,v
retrieving revision 1.1
retrieving revision 1.2
diff -C2 -d -r1.1 -r1.2
*** mapilib.i 1 Sep 1999 23:17:16 -0000 1.1
--- mapilib.i 1 Sep 2003 08:05:36 -0000 1.2
***************
*** 79,82 ****
--- 79,86 ----
$target = &temp;
}
+ %typemap(python,ignore) IStream **OUTPUT(IStream *temp)
+ {
+ $target = &temp;
+ }
***************
*** 120,123 ****
--- 124,130 ----
MAKE_OUTPUT_INTERFACE($source, $target, IID_IMsgServiceAdmin)
}
+ %typemap(python,argout) IStream **OUTPUT {
+ MAKE_OUTPUT_INTERFACE($source, $target, IID_IStream)
+ }
***************
*** 147,151 ****
IAddrBook *INPUT_NULLOK,
IMsgServiceAdmin *INPUT,
! IMsgServiceAdmin *INPUT_NULLOK
{
if ($source) $source->Release();
--- 154,160 ----
IAddrBook *INPUT_NULLOK,
IMsgServiceAdmin *INPUT,
! IMsgServiceAdmin *INPUT_NULLOK,
! IStream *INPUT,
! IStream *INPUT_NULLOK
{
if ($source) $source->Release();
***************
*** 267,270 ****
--- 276,288 ----
return NULL;
}
+ %typemap(python,in) IStream *INPUT {
+ if (!PyCom_InterfaceFromPyInstanceOrObject($source, IID_IStream, (void **)&$target, 0))
+ return NULL;
+ }
+ %typemap(python,in) IStream *INPUT_NULLOK {
+ if (!PyCom_InterfaceFromPyInstanceOrObject($source, IID_IStream, (void **)&$target, 1))
+ return NULL;
+ }
+
// Some ** special cases.
%typemap(python,freearg) IMsgStore **INPUT
|
