From 4e331a687818d56979c456c25a7ece4a4128171e Mon Sep 17 00:00:00 2001
From: "Mirar (Pontus Hagland)" <pike@sort.mirar.org>
Date: Sat, 12 Jun 1999 21:13:08 +0200
Subject: [PATCH] it works, abit
Rev: src/modules/Parser/html.c:1.18
Rev: src/modules/Parser/module.pmod.in:1.2
---
src/modules/Parser/html.c | 150 ++++++++++++++++++++++++--
src/modules/Parser/module.pmod.in | 169 +++++++++++++++++++++++++++++-
2 files changed, 310 insertions(+), 9 deletions(-)
diff --git a/src/modules/Parser/html.c b/src/modules/Parser/html.c
index c3de86ac65..131d9b6431 100644
--- a/src/modules/Parser/html.c
+++ b/src/modules/Parser/html.c
@@ -19,8 +19,8 @@
extern struct program *parser_html_program;
-#define DEBUG
/*
+#define DEBUG
#define SCAN_DEBUG
*/
@@ -33,7 +33,7 @@ extern struct program *parser_html_program;
#define DEBUG_MARK_SPOT(TEXT,FEED,C) do; while(0)
#endif
-#if 1
+#if 0
#define free(X) fprintf(stderr,"free line %d: %p\n",__LINE__,X); free(X)
#endif
@@ -465,13 +465,14 @@ static void html__set_entity_callback(INT32 args)
**! (Ie, do destructive editing of the args mapping if you
**! want the next callback to read it - don't just return a new tag.)
**! </ul>
+**!
+**! see also: tags, containers, entities
**!
*/
static void html_add_tag(INT32 args)
{
- check_all_args("add_tag",args,
- BIT_STRING,BIT_MIXED,0);
+ check_all_args("add_tag",args,BIT_STRING,BIT_MIXED,0);
if (THIS->maptag->refs>1)
{
push_mapping(THIS->maptag);
@@ -486,8 +487,7 @@ static void html_add_tag(INT32 args)
static void html_add_container(INT32 args)
{
- check_all_args("add_container",args,
- BIT_STRING,BIT_MIXED,0);
+ check_all_args("add_container",args,BIT_STRING,BIT_MIXED,0);
if (THIS->mapcont->refs>1)
{
push_mapping(THIS->mapcont);
@@ -500,8 +500,7 @@ static void html_add_container(INT32 args)
static void html_add_entity(INT32 args)
{
- check_all_args("parse_tag_args",args,
- BIT_STRING,BIT_MIXED,0);
+ check_all_args("add_entity",args,BIT_STRING,BIT_MIXED,0);
if (THIS->mapentity->refs>1)
{
push_mapping(THIS->mapentity);
@@ -512,6 +511,81 @@ static void html_add_entity(INT32 args)
pop_n_elems(args);
}
+static void html_add_tags(INT32 args)
+{
+ int sz;
+ INT32 e;
+ struct keypair *k;
+ check_all_args("add_tags",args,BIT_MAPPING,0);
+
+ sz=sp[-1].u.mapping->size;
+
+ MAPPING_LOOP(sp[-1].u.mapping)
+ {
+ push_svalue(&k->ind);
+ push_svalue(&k->val);
+ html_add_tag(2);
+ }
+
+ pop_n_elems(args);
+}
+
+static void html_add_containers(INT32 args)
+{
+ int sz;
+ INT32 e;
+ struct keypair *k;
+ check_all_args("add_containers",args,BIT_MAPPING,0);
+
+ sz=sp[-1].u.mapping->size;
+
+ MAPPING_LOOP(sp[-1].u.mapping)
+ {
+ push_svalue(&k->ind);
+ push_svalue(&k->val);
+ html_add_container(2);
+ }
+
+ pop_n_elems(args);
+}
+
+static void html_add_entities(INT32 args)
+{
+ int sz;
+ INT32 e;
+ struct keypair *k;
+ check_all_args("add_entities",args,BIT_MAPPING,0);
+
+ sz=sp[-1].u.mapping->size;
+
+ MAPPING_LOOP(sp[-1].u.mapping)
+ {
+ push_svalue(&k->ind);
+ push_svalue(&k->val);
+ html_add_entity(2);
+ }
+
+ pop_n_elems(args);
+}
+
+static void html_tags(INT32 args)
+{
+ pop_n_elems(args);
+ ref_push_mapping(THIS->maptag);
+}
+
+static void html_containers(INT32 args)
+{
+ pop_n_elems(args);
+ ref_push_mapping(THIS->mapcont);
+}
+
+static void html_entities(INT32 args)
+{
+ pop_n_elems(args);
+ ref_push_mapping(THIS->mapentity);
+}
+
/****** try_feed - internal main ********************/
/* ---------------------------------------- */
@@ -1922,6 +1996,43 @@ void html_write_out(INT32 args)
/** query *******************************************/
+/*
+**! method array(int) at();
+**! method int at_line();
+**! method int at_char();
+**! method int at_column();
+**! Get the current position.
+**! <ref>at</ref>() gives an array consisting of
+**! ({<i>line</i>,<i>char</i>,<i>column</i>}), in that order.
+*/
+
+static void html_at_line(INT32 args)
+{
+ pop_n_elems(args);
+ push_int(THIS->stack->pos.lineno);
+}
+
+static void html_at_char(INT32 args)
+{
+ pop_n_elems(args);
+ push_int(THIS->stack->pos.byteno);
+}
+
+static void html_at_column(INT32 args)
+{
+ pop_n_elems(args);
+ push_int(THIS->stack->pos.byteno-THIS->stack->pos.linestart);
+}
+
+static void html_at(INT32 args)
+{
+ pop_n_elems(args);
+ push_int(THIS->stack->pos.lineno);
+ push_int(THIS->stack->pos.byteno);
+ push_int(THIS->stack->pos.byteno-THIS->stack->pos.linestart);
+ f_aggregate(3);
+}
+
/*
**! method string current()
**! Gives the current range of data, ie the contents
@@ -2182,6 +2293,8 @@ void html__inspect(INT32 args)
f_aggregate_mapping(n*2);
}
+/** create, clone ***********************************/
+
void html_create(INT32 args)
{
pop_n_elems(args);
@@ -2292,6 +2405,12 @@ void init_parser_html(void)
/* query */
ADD_FUNCTION("current",html_current,tFunc(,tStr),0);
+
+ ADD_FUNCTION("at",html_at,tFunc(,tArr(tInt)),0);
+ ADD_FUNCTION("at_line",html_at_line,tFunc(,tInt),0);
+ ADD_FUNCTION("at_char",html_at_char,tFunc(,tInt),0);
+ ADD_FUNCTION("at_column",html_at_column,tFunc(,tInt),0);
+
ADD_FUNCTION("tag_name",html_tag_name,tFunc(,tStr),0);
ADD_FUNCTION("tag_args",html_tag_args,tFunc(,tMapping),0);
@@ -2304,6 +2423,21 @@ void init_parser_html(void)
ADD_FUNCTION("add_entity",html_add_entity,
tFunc(tStr tTodo(""),tVoid),0);
+ ADD_FUNCTION("add_tags",html_add_tags,
+ tFunc(tMap(tStr,tTodo( tTagargs )),tVoid),0);
+ ADD_FUNCTION("add_containers",html_add_containers,
+ tFunc(tMap(tStr,tTodo( tTagargs tStr )),tVoid),0);
+ ADD_FUNCTION("add_entities",html_add_entities,
+ tFunc(tMap(tStr,tTodo( "" )),tVoid),0);
+
+ ADD_FUNCTION("tags",html_tags,
+ tFunc(,tMap(tStr,tTodo( tTagargs ))),0);
+ ADD_FUNCTION("containers",html_containers,
+ tFunc(,tMap(tStr,tTodo( tTagargs tStr ))),0);
+ ADD_FUNCTION("entities",html_entities,
+ tFunc(,tMap(tStr,tTodo( "" ))),0);
+
+
/* special callbacks */
ADD_FUNCTION("_set_tag_callback",html__set_tag_callback,
diff --git a/src/modules/Parser/module.pmod.in b/src/modules/Parser/module.pmod.in
index 747882d115..3089d84fe4 100644
--- a/src/modules/Parser/module.pmod.in
+++ b/src/modules/Parser/module.pmod.in
@@ -1,8 +1,175 @@
/*
- * $Id: module.pmod.in,v 1.1 1999/02/19 04:58:40 mirar Exp $
+ * $Id: module.pmod.in,v 1.2 1999/06/12 19:13:08 mirar Exp $
*
*/
inherit @module@;
+//! module Parser
+
+class SGML
+//! class SGML
+//! This is a handy simple parser of SGML-like
+//! syntax like HTML. It doesn't do anything advanced,
+//! but finding the corresponding end-tags.
+//!
+//! It's used like this:
+//! <pre>array res=Parser.SGML()->feed(string)->finish()->result();</pre>
+//!
+//! The resulting structure is an array of atoms,
+//! where the atom can be a string or a tag.
+//! A tag contains a similar array, as data.
+//!
+//! Example:
+//! A string
+//! <tt>"<gat> <gurka> </gurka> <banan> <kiwi> </gat>"</tt>
+//! results in
+//! <pre>
+//! ({
+//! tag "gat" object with data:
+//! ({
+//! tag "gurka" object with data:
+//! ({
+//! " "
+//! })
+//! tag "banan" object with data:
+//! ({
+//! " "
+//! tag "kiwi" object with data:
+//! ({
+//! " "
+//! })
+//! })
+//! })
+//! })
+//! </pre>
+//!
+//! ie, simple "tags" (not containers) are not detected,
+//! but containers are ended implicitely by a surrounding
+//! container _with_ an end tag.
+//!
+//! The 'tag' is an object with the following variables:
+//! <pre>
+//! string name; - name of tag
+//! mapping args; - argument to tag
+//! int line,char,column; - position of tag
+//! string file; - filename (see <ref>create</ref>)
+//! array(SGMLatom) data; - contained data
+//! </pre>
+//!
+{
+ string file;
+
+ class SGMLatom
+ {
+ string name;
+ mapping args;
+ int line,char,column;
+ string file;
+ array(SGMLatom) data;
+ }
+
+ static array(array(object(SGMLatom)|string)) res=({({})});
+ static array(SGMLatom) tagstack=({});
+ static array(object) errors;
+
+ array(object(SGMLatom)|string) data;
+
+ static private array(string) got_tag(object g)
+ {
+ string name=g->tag_name();
+
+ if (name!="" && name[0]=='/')
+ {
+ int i=search(tagstack->name,name[1..]);
+ if (i!=-1)
+ {
+ i++;
+ while (i--)
+ {
+ tagstack[0]->data=res[0];
+ res=res[1..];
+ tagstack=tagstack[1..];
+ }
+ return ({});
+ }
+ }
+
+ object t=SGMLatom();
+ t->name=name;
+ t->args=g->tag_args();
+ [t->line,t->char,t->column]=g->at();
+ t->file=file;
+ res[0]+=({t});
+ tagstack=({t})+tagstack;
+ res=({({})})+res;
+
+ return ({}); // don't care
+ }
+
+ void debug(array|void arr,void|int level)
+ {
+ level+=2;
+ if (!arr) arr=data;
+ foreach (arr,string|object(SGMLatom) t)
+ if (stringp(t))
+ write("%*s%-=*s\n",level,"",79-level,sprintf("%O",t));
+ else
+ {
+ write("%*stag %O\n",level,"",t->name,);
+ if (sizeof(t->args))
+ write("%*s%-=*s\n",level+4,"",75-level,sprintf("%O",t->args));
+ debug(t->data,level);
+ }
+ }
+
+
+ private static object p=HTML();
+
+//! static void create()
+//! static void create(string filename)
+//! This object is created with this filename.
+//! It's passed to all created tags, for debug and trace purposes.
+//! note:
+//! No, it doesn't read the file itself. See <ref>feed</ref>.
+
+ void create(void|string _file)
+ {
+ file=_file;
+
+ p->_set_tag_callback(got_tag);
+ p->_set_data_callback(lambda(object g,string data)
+ { if (data!="") res[0]+=({data}); return ({}); });
+ }
+
+//! static object feed(string s)
+//! static array finish()
+//! static array result(string s)
+//! Feed new data to the object, or finish the stream.
+//! No result can be used until finish() is called.
+//!
+//! Both finish() and result() returns the computed data.
+//!
+//! feed() returns the called object.
+
+ object feed(string s)
+ {
+ p->feed(s);
+ return this_object();
+ }
+
+ array finish()
+ {
+ p->finish();
+ data=res[0];
+ res=0;
+ return data;
+ }
+
+ array(object(SGMLatom)|string) result()
+ {
+ return data;
+ }
+}
+
--
GitLab