path: root/camel/camel-mime-part-utils.c

/* -*- Mode: C; tab-width: 8; indent-tabs-mode: t; c-basic-offset: 8; fill-column: 160 -*- */
/* camel-mime-part-utils : Utility for mime parsing and so on
 *
 * Authors: Bertrand Guiheneuf <bertrand@helixcode.com>
 *          Michael Zucchi <notzed@ximian.com>
 *          Jeffrey Stedfast <fejj@ximian.com>
 *
 * Copyright 1999, 2000 Ximian, Inc. (www.ximian.com)
 *
 * This program is free software; you can redistribute it and/or 
 * modify it under the terms of version 2 of the GNU General Public 
 * License as published by the Free Software Foundation.
 *
 * This program is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 * GNU General Public License for more details.
 *
 * You should have received a copy of the GNU General Public License
 * along with this program; if not, write to the Free Software
 * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307
 * USA
 */

#ifdef HAVE_CONFIG_H
#include <config.h>
#endif

#include <stdio.h>
#include <string.h>
#include <unistd.h>
#include <ctype.h>
#include <errno.h>

#include <gal/util/e-iconv.h>

#include "string-utils.h"
#include "camel-charset-map.h"
#include "camel-mime-part-utils.h"
#include "camel-mime-message.h"
#include "camel-multipart.h"
#include "camel-multipart-signed.h"
#include "camel-multipart-encrypted.h"
#include "camel-seekable-substream.h"
#include "camel-stream-fs.h"
#include "camel-stream-filter.h"
#include "camel-stream-mem.h"
#include "camel-mime-filter-basic.h"
#include "camel-mime-filter-charset.h"
#include "camel-mime-filter-crlf.h"
#include "camel-mime-filter-save.h"
#include "camel-html-parser.h"

#define d(x) /*(printf("%s(%d): ", __FILE__, __LINE__),(x))
           #include <stdio.h>*/

/* example: <META http-equiv="Content-Type" content="text/html; charset=ISO-8859-1"> */

static const char *
check_html_charset(char *buffer, int length)
{
    CamelHTMLParser *hp;
    const char *charset = NULL;
    camel_html_parser_t state;
    struct _header_content_type *ct;

    /* if we need to first base64/qp decode, do this here, sigh */
    hp = camel_html_parser_new();
    camel_html_parser_set_data(hp, buffer, length, TRUE);
    
    do {
        const char *data;
        int len;
        const char *val;
        
        state = camel_html_parser_step(hp, &data, &len);
        
        /* example: <META http-equiv="Content-Type" content="text/html; charset=ISO-8859-1"> */
        
        switch(state) {
        case CAMEL_HTML_PARSER_ELEMENT:
            val = camel_html_parser_tag(hp);
            d(printf("Got tag: %s\n", val));
            if (strcasecmp(val, "meta") == 0
                && (val = camel_html_parser_attr(hp, "http-equiv"))
                && strcasecmp(val, "content-type") == 0
                && (val = camel_html_parser_attr(hp, "content"))
                && (ct = header_content_type_decode(val))) {
                charset = header_content_type_param(ct, "charset");
                charset = e_iconv_charset_name (charset);
                header_content_type_unref(ct);
            }
            break;
        default:
            /* ignore everything else */
            break;
        }
    } while (charset == NULL && state != CAMEL_HTML_PARSER_EOF);

    camel_object_unref (hp);

    return charset;
}

static GByteArray *
convert_buffer (GByteArray *in, const char *to, const char *from)
{
    size_t inleft, outleft, outlen, converted = 0;
    GByteArray *out = NULL;
    const char *inbuf;
    char *outbuf;
    iconv_t cd;
    
    if (in->len == 0)
        return g_byte_array_new();
    
    d(printf("converting buffer from %s to %s:\n", from, to));
    d(fwrite(in->data, 1, (int)in->len, stdout));
    d(printf("\n"));
    
    cd = e_iconv_open(to, from);
    if (cd == (iconv_t) -1) {
        g_warning ("Cannot convert from '%s' to '%s': %s", from, to, strerror (errno));
        return NULL;
    }
    
    outlen = in->len * 2 + 16;
    out = g_byte_array_new ();
    g_byte_array_set_size (out, outlen);
    
    inbuf = in->data;
    inleft = in->len;
    
    do {
        outbuf = out->data + converted;
        outleft = outlen - converted;
        
        converted = e_iconv (cd, &inbuf, &inleft, &outbuf, &outleft);
        if (converted == (size_t) -1) {
            if (errno != E2BIG && errno != EINVAL)
                goto fail;
        }
        
        /*
         * E2BIG   There is not sufficient room at *outbuf.
         *
         * We just need to grow our outbuffer and try again.
         */
        
        converted = outbuf - (char *)out->data;
        if (errno == E2BIG) {
            outlen += inleft * 2 + 16;
            out = g_byte_array_set_size (out, outlen);
            outbuf = out->data + converted;
        }
        
    } while (errno == E2BIG && inleft > 0);
    
    /*
     * EINVAL  An  incomplete  multibyte sequence has been encoun­
     *         tered in the input.
     *
     * We'll just have to ignore it...
     */
    
    /* flush the iconv conversion */
    e_iconv (cd, NULL, NULL, &outbuf, &outleft);
    
    /* now set the true length on the GByteArray */
    converted = outbuf - (char *)out->data;
    g_byte_array_set_size (out, converted);
    
    d(printf("converted data:\n"));
    d(fwrite(out->data, 1, (int)out->len, stdout));
    d(printf("\n"));
    
    e_iconv_close (cd);
    
    return out;
    
 fail:
    g_warning ("Cannot convert from '%s' to '%s': %s", from, to, strerror (errno));
    
    g_byte_array_free (out, TRUE);
    
    e_iconv_close (cd);
    
    return NULL;
}

/* We don't really use the charset argument except for debugging... */
static gboolean
broken_windows_charset (GByteArray *buffer, const char *charset)
{
    register unsigned char *inptr;
    unsigned char *inend;
    
    inptr = buffer->data;
    inend = inptr + buffer->len;
    
    while (inptr < inend) {
        register unsigned char c = *inptr++;
        
        if (c >= 128 && c <= 159) {
            g_warning ("Encountered Windows charset parading as %s", charset);
            return TRUE;
        }
    }
    
    return FALSE;
}

static gboolean
is_7bit (GByteArray *buffer)
{
    register unsigned int i;
    
    for (i = 0; i < buffer->len; i++)
        if (buffer->data[i] > 127)
            return FALSE;
    
    return TRUE;
}

static const char *iso_charsets[] = {
    "us-ascii",
    "iso-8859-1",
    "iso-8859-2",
    "iso-8859-3",
    "iso-8859-4",
    "iso-8859-5",
    "iso-8859-6",
    "iso-8859-7",
    "iso-8859-8",
    "iso-8859-9",
    "iso-8859-10",
    "iso-8859-11",
    "iso-8859-12",
    "iso-8859-13",
    "iso-8859-14",
    "iso-8859-15",
    "iso-8859-16"
};

#define NUM_ISO_CHARSETS (sizeof (iso_charsets) / sizeof (iso_charsets[0]))

static const char *
canon_charset_name (const char *charset)
{
    const char *ptr;
    char *endptr;
    int iso;
    
    if (strncasecmp (charset, "iso", 3) != 0)
        return charset;
    
    ptr = charset + 3;
    if (*ptr == '-' || *ptr == '_')
        ptr++;
    
    /* if it's not an iso-8859-# charset, we don't care about it */
    if (strncmp (ptr, "8859", 4) != 0)
        return charset;
    
    ptr += 4;
    if (*ptr == '-' || *ptr == '_')
        ptr++;
    
    iso = strtoul (ptr, &endptr, 10);
    if (endptr == ptr || *endptr != '\0')
        return charset;
    
    if (iso >= NUM_ISO_CHARSETS)
        return charset;
    
    return iso_charsets[iso];
}

/* simple data wrapper */
static void
simple_data_wrapper_construct_from_parser (CamelDataWrapper *dw, CamelMimeParser *mp)
{
    CamelMimeFilter *fdec = NULL, *fcrlf = NULL;
    CamelMimeFilterBasicType enctype = 0;
    size_t len;
    int decid = -1, crlfid = -1;
    struct _header_content_type *ct;
    const char *charset = NULL;
    char *encoding, *buf;
    GByteArray *buffer;
    CamelStream *mem;
    
    d(printf ("simple_data_wrapper_construct_from_parser()\n"));
    
    /* first, work out conversion, if any, required, we dont care about what we dont know about */
    encoding = header_content_encoding_decode (camel_mime_parser_header (mp, "Content-Transfer-Encoding", NULL));
    if (encoding) {
        if (!strcasecmp (encoding, "base64")) {
            d(printf("Adding base64 decoder ...\n"));
            enctype = CAMEL_MIME_FILTER_BASIC_BASE64_DEC;
        } else if (!strcasecmp (encoding, "quoted-printable")) {
            d(printf("Adding quoted-printable decoder ...\n"));
            enctype = CAMEL_MIME_FILTER_BASIC_QP_DEC;
        } else if (!strcasecmp (encoding, "x-uuencode")) {
            d(printf("Adding uudecoder ...\n"));
            enctype = CAMEL_MIME_FILTER_BASIC_UU_DEC;
        }
        g_free (encoding);
        
        if (enctype != 0) {
            fdec = (CamelMimeFilter *)camel_mime_filter_basic_new_type(enctype);
            decid = camel_mime_parser_filter_add (mp, fdec);
        }
    }
    
    /* If we're doing text, we also need to do CRLF->LF and may have to convert it to UTF8 as well. */
    ct = camel_mime_parser_content_type (mp);
    if (header_content_type_is (ct, "text", "*")) {
        charset = header_content_type_param (ct, "charset");
        charset = e_iconv_charset_name (charset);
        
        if (fdec) {
            d(printf ("Adding CRLF conversion filter\n"));
            fcrlf = camel_mime_filter_crlf_new (CAMEL_MIME_FILTER_CRLF_DECODE,
                                CAMEL_MIME_FILTER_CRLF_MODE_CRLF_ONLY);
            crlfid = camel_mime_parser_filter_add (mp, fcrlf);
        }
    }
    
    /* read in the entire content */
    buffer = g_byte_array_new ();
    while (camel_mime_parser_step (mp, &buf, &len) != HSCAN_BODY_END) {
        d(printf("appending o/p data: %d: %.*s\n", len, len, buf));
        g_byte_array_append (buffer, buf, len);
    }
    
    /* check for broken Outlook/Web mailers that like to send html marked as text/plain */
    if (header_content_type_is (ct, "text", "plain")) {
        register const unsigned char *inptr;
        const unsigned char *inend;
        
        inptr = buffer->data;
        inend = inptr + buffer->len;
        
        while (inptr < inend && isspace ((int) *inptr))
            inptr++;

        if (((inend-inptr) > 5 && g_ascii_strncasecmp(inptr, "<html", 5) == 0)
            || ((inend-inptr) > 9 && g_ascii_strncasecmp(inptr, "<!doctype", 9) == 0)) {
            /* re-tag as text/html */
            g_free (ct->subtype);
            ct->subtype = g_strdup ("html");
        }
    }
    
    /* Possible Lame Mailer Alert... check the META tags for a charset */
    if (!charset && header_content_type_is (ct, "text", "html")) {
        if ((charset = check_html_charset (buffer->data, buffer->len)))
            header_content_type_set_param (ct, "charset", charset);
    }
    
    /* if we need to do charset conversion, see if we can/it works/etc */
    if (charset && !(strcasecmp (charset, "us-ascii") == 0
             || strcasecmp (charset, "utf-8") == 0
             || strncasecmp (charset, "x-", 2) == 0)) {
        GByteArray *out;
        
        /* You often see Microsoft Windows users announcing their texts
         * as being in ISO-8859-1 even when in fact they contain funny
         * characters from the Windows-CP1252 superset.
         */
        charset = canon_charset_name (charset);
        if (!strncasecmp (charset, "iso-8859", 8)) {
            /* check for Windows-specific chars... */
            if (broken_windows_charset (buffer, charset))
                charset = camel_charset_iso_to_windows (charset);
        }
        
        out = convert_buffer (buffer, "UTF-8", charset);
        if (out) {
            /* converted ok, use this data instead */
            g_byte_array_free(buffer, TRUE);
            buffer = out;
        } else {
            /* else failed to convert, leave as raw? */
            g_warning("Storing text as raw, unknown charset '%s' or invalid format", charset);
            dw->rawtext = TRUE;
        }
    } else if (header_content_type_is (ct, "text", "*")) {
        if (charset == NULL || !strcasecmp (charset, "us-ascii")) {
            /* check that it's 7bit */
            dw->rawtext = !is_7bit (buffer);
        } else if (!strncasecmp (charset, "x-", 2)) {
            /* we're not even going to bother trying to convert, so set the
               rawtext bit to TRUE and let the mailer deal with it. */
            dw->rawtext = TRUE;
        } else if (!strcasecmp (charset, "utf-8") && buffer->len) {
            /* check that it is valid utf8 */
            dw->rawtext = !g_utf8_validate (buffer->data, buffer->len, NULL);
        }
    }
    
    d(printf("message part kept in memory!\n"));
    
    mem = camel_stream_mem_new_with_byte_array(buffer);
    camel_data_wrapper_construct_from_stream(dw, mem);
    camel_object_unref((CamelObject *)mem);

    camel_mime_parser_filter_remove(mp, decid);
    camel_mime_parser_filter_remove(mp, crlfid);
    
    if (fdec)
        camel_object_unref((CamelObject *)fdec);
    if (fcrlf)
        camel_object_unref((CamelObject *)fcrlf);
}

/* This replaces the data wrapper repository ... and/or could be replaced by it? */
void
camel_mime_part_construct_content_from_parser (CamelMimePart *dw, CamelMimeParser *mp)
{
    CamelDataWrapper *content = NULL;
    CamelContentType *ct;
    
    ct = camel_mime_parser_content_type (mp);

    switch (camel_mime_parser_state (mp)) {
    case HSCAN_HEADER:
        d(printf("Creating body part\n"));
        /* multipart/signed is some fucked up type that we must treat as binary data, fun huh, idiots. */
        if (header_content_type_is (ct, "multipart", "signed")) {
            content = (CamelDataWrapper *) camel_multipart_signed_new ();
            camel_multipart_construct_from_parser ((CamelMultipart *) content, mp);
        } else {
            content = camel_data_wrapper_new ();
            simple_data_wrapper_construct_from_parser (content, mp);
        }
        break;
    case HSCAN_MESSAGE:
        d(printf("Creating message part\n"));
        content = (CamelDataWrapper *) camel_mime_message_new ();
        camel_mime_part_construct_from_parser ((CamelMimePart *)content, mp);
        break;
    case HSCAN_MULTIPART:
        d(printf("Creating multi-part\n"));
        if (header_content_type_is (ct, "multipart", "encrypted"))
            content = (CamelDataWrapper *) camel_multipart_encrypted_new ();
        else if (header_content_type_is (ct, "multipart", "signed"))
            content = (CamelDataWrapper *) camel_multipart_signed_new ();
        else
            content = (CamelDataWrapper *) camel_multipart_new ();
        
        camel_multipart_construct_from_parser((CamelMultipart *)content, mp);
        d(printf("Created multi-part\n"));
        break;
    default:
        g_warning("Invalid state encountered???: %d", camel_mime_parser_state (mp));
    }
    if (content) {
        /* would you believe you have to set this BEFORE you set the content object???  oh my god !!!! */
        camel_data_wrapper_set_mime_type_field (content, camel_mime_part_get_content_type (dw));
        camel_medium_set_content_object ((CamelMedium *)dw, content);
        
        /* Note: we don't set ct as the content-object's mime-type above because
         * camel_medium_set_content_object() may re-write the Content-Type header
         * (see CamelMimePart::set_content_object) if we did that (which is a Bad Thing).
         * However, if we set it *afterward*, we can still use any special auto-detections
         * that we found in simple_data_wrapper_construct_from_parser(). This is important
         * later when we go to render the MIME parts in mail-format.c */
        camel_data_wrapper_set_mime_type_field (content, ct);
        
        camel_object_unref (content);
    }
}