X-Git-Url: http://git.indexdata.com/?a=blobdiff_plain;f=retrieval%2Fd1_read.c;h=7f3c4a3265e3664b0d2c64d77884a67c461595ee;hb=424b34cfd7042e0c854ba2ca1f2cc656b61f3e46;hp=66baf41cc3a20db3a2decf2dea20c57fcb0bd1b1;hpb=a774858896a9b56c966edbfb70d8eae1c91c4ad1;p=yaz-moved-to-github.git diff --git a/retrieval/d1_read.c b/retrieval/d1_read.c index 66baf41..7f3c4a3 100644 --- a/retrieval/d1_read.c +++ b/retrieval/d1_read.c @@ -3,13 +3,19 @@ * See the file LICENSE for details. * Sebastian Hammer, Adam Dickmeiss * - * $Id: d1_read.c,v 1.44 2002-07-03 10:04:04 adam Exp $ + * $Id: d1_read.c,v 1.48 2002-08-19 21:09:10 adam Exp $ */ #include #include #include +#include + +#if HAVE_ICONV_H +#include +#endif + #include #include #include @@ -18,9 +24,12 @@ data1_node *data1_get_root_tag (data1_handle dh, data1_node *n) { if (!n) return 0; - n = n->child; - while (n && n->which != DATA1N_tag) - n = n->next; + if (data1_is_xmlmode(dh)) + { + n = n->child; + while (n && n->which != DATA1N_tag) + n = n->next; + } return n; } @@ -30,9 +39,19 @@ data1_node *data1_get_root_tag (data1_handle dh, data1_node *n) */ data1_node *get_parent_tag (data1_handle dh, data1_node *n) { - for (; n && n->which != DATA1N_root; n = n->parent) - if (n->which == DATA1N_tag) - return n; + if (data1_is_xmlmode(dh)) + { + for (; n && n->which != DATA1N_root; n = n->parent) + if (n->which == DATA1N_tag && n->parent && + n->parent->which != DATA1N_root) + return n; + } + else + { + for (; n && n->which != DATA1N_root; n = n->parent) + if (n->which == DATA1N_tag) + return n; + } return 0; } @@ -176,16 +195,24 @@ data1_node *data1_mk_tag_n (data1_handle dh, NMEM nmem, { data1_node *partag = get_parent_tag(dh, at); data1_node *res = data1_mk_node2 (dh, nmem, DATA1N_tag, at); - data1_element *e = NULL; data1_xattr **p; + data1_element *e = 0; res->u.tag.tag = data1_insert_string_n (dh, res, nmem, tag, len); - if (partag) - e = partag->u.tag.element; - res->u.tag.element = - data1_getelementbytagname (dh, at->root->u.root.absyn, - e, res->u.tag.tag); + if (!partag) /* top tag? */ + e = data1_getelementbytagname (dh, at->root->u.root.absyn, + 0 /* index as local */, + res->u.tag.tag); + else + { + /* only set element for known tags */ + e = partag->u.tag.element; + if (e) + e = data1_getelementbytagname (dh, at->root->u.root.absyn, + e, res->u.tag.tag); + } + res->u.tag.element = e; p = &res->u.tag.attributes; while (attr && *attr) { @@ -329,9 +356,14 @@ static data1_node *data1_add_insert_taggeddata(data1_handle dh, data1_node *datn = 0; data1_node *tagn = 0; - if (partag) + if (!partag) + e = data1_getelementbytagname (dh, root->u.root.absyn, 0, tagname); + else + { e = partag->u.tag.element; - e = data1_getelementbytagname (dh, root->u.root.absyn, e, tagname); + if (e) + e = data1_getelementbytagname (dh, root->u.root.absyn, e, tagname); + } if (local_allowed || e) { tagn = data1_mk_node2 (dh, m, DATA1N_tag, at); @@ -616,8 +648,16 @@ data1_node *data1_read_nodex (data1_handle dh, NMEM m, break; } } - if (level <= 1) - return d1_stack[0]; + if (data1_is_xmlmode(dh)) + { + if (level <= 1) + return d1_stack[0]; + } + else + { + if (level <= 0) + return d1_stack[0]; + } continue; } else if (!strcmp(tag, "var")) @@ -670,10 +710,20 @@ data1_node *data1_read_nodex (data1_handle dh, NMEM m, if (level == 0) { parent = data1_mk_root (dh, m, tag); - d1_stack[level++] = parent; + res = d1_stack[level] = parent; + + if (data1_is_xmlmode(dh)) + { + level++; + res = data1_mk_tag (dh, m, tag, 0 /* attr */, parent); + res->u.tag.attributes = xattr; + } + } + else + { + res = data1_mk_tag (dh, m, tag, 0 /* attr */, parent); + res->u.tag.attributes = xattr; } - res = data1_mk_tag (dh, m, tag, 0 /* attr */, parent); - res->u.tag.attributes = xattr; } d1_stack[level] = res; d1_stack[level+1] = 0; @@ -793,3 +843,131 @@ data1_node *data1_read_sgml (data1_handle dh, NMEM m, const char *buf) return data1_read_node (dh, &bp, m); } + +#if HAVE_ICONV_H + +static int conv_item (NMEM m, iconv_t t, + WRBUF wrbuf, char *inbuf, size_t inlen) +{ + wrbuf_rewind (wrbuf); + if (wrbuf->size < 10) + wrbuf_grow (wrbuf, 10); + for (;;) + { + char *outbuf = wrbuf->buf + wrbuf->pos; + size_t outlen = wrbuf->size - wrbuf->pos; + if (iconv (t, &inbuf, &inlen, &outbuf, &outlen) == + (size_t)(-1) && errno != E2BIG) + { + /* bad data. stop and skip conversion entirely */ + return -1; + } + else if (inlen == 0) + { /* finished converting */ + wrbuf->pos = wrbuf->size - outlen; + break; + } + else + { + /* buffer too small: make sure we expand buffer */ + wrbuf->pos = wrbuf->size - outlen; + wrbuf_grow(wrbuf, 20); + } + } + return 0; +} + +static void data1_iconv_s (data1_handle dh, NMEM m, data1_node *n, + iconv_t t, WRBUF wrbuf, const char *tocode) +{ + for (; n; n = n->next) + { + switch (n->which) + { + case DATA1N_data: + case DATA1N_comment: + if (conv_item (m, t, wrbuf, n->u.data.data, n->u.data.len) == 0) + { + n->u.data.data = + data1_insert_string_n (dh, n, m, wrbuf->buf, + wrbuf->pos); + n->u.data.len = wrbuf->pos; + } + break; + case DATA1N_tag: + if (conv_item (m, t, wrbuf, n->u.tag.tag, strlen(n->u.tag.tag)) + == 0) + { + n->u.tag.tag = + data1_insert_string_n (dh, n, m, + wrbuf->buf, wrbuf->pos); + } + if (n->u.tag.attributes) + { + data1_xattr *p; + for (p = n->u.tag.attributes; p; p = p->next) + { + if (p->value && + conv_item(m, t, wrbuf, p->value, strlen(p->value)) + == 0) + { + wrbuf_puts (wrbuf, ""); + p->value = nmem_strdup (m, wrbuf->buf); + } + } + } + break; + case DATA1N_preprocess: + if (strcmp(n->u.preprocess.target, "xml") == 0) + { + data1_xattr *p = n->u.preprocess.attributes; + for (; p; p = p->next) + if (strcmp (p->name, "encoding") == 0) + p->value = nmem_strdup (m, tocode); + } + break; + } + data1_iconv_s (dh, m, n->child, t, wrbuf, tocode); + } +} +#endif + +const char *data1_get_encoding (data1_handle dh, data1_node *n) +{ + /* see if we have an xml header that specifies encoding */ + if (n && n->child && n->child->which == DATA1N_preprocess && + strcmp (n->child->u.preprocess.target, "xml") == 0) + { + data1_xattr *xp = n->child->u.preprocess.attributes; + for (; xp; xp = xp->next) + if (!strcmp (xp->name, "encoding") == 0) + return xp->value; + } + /* no encoding in header, so see if "encoding" was specified for abs */ + if (n && n->which == DATA1N_root && + n->u.root.absyn && n->u.root.absyn->encoding) + return n->u.root.absyn->encoding; + /* none of above, return a hard coded default */ + return "ISO-8859-1"; +} + +int data1_iconv (data1_handle dh, NMEM m, data1_node *n, + const char *tocode, + const char *fromcode) +{ +#if HAVE_ICONV_H + if (strcmp (tocode, fromcode)) + { + WRBUF wrbuf = wrbuf_alloc(); + iconv_t t = iconv_open (tocode, fromcode); + if (t == (iconv_t) (-1)) + return -1; + data1_iconv_s (dh, m, n, t, wrbuf, tocode); + iconv_close (t); + wrbuf_free (wrbuf, 1); + } + return 0; +#else + return -2; +#endif +}