From 660b569b0980fc8f71b03ed666dd02eec8388b4c Mon Sep 17 00:00:00 2001 From: Xavier Roche Date: Mon, 19 Mar 2012 12:59:03 +0000 Subject: httrack 3.41.2 --- html/plug.html | 256 +++++++++++++++++++++++++++++++++++++++++++-------------- 1 file changed, 193 insertions(+), 63 deletions(-) (limited to 'html/plug.html') diff --git a/html/plug.html b/html/plug.html index 42b0895..1b82c46 100755 --- a/html/plug.html +++ b/html/plug.html @@ -108,85 +108,215 @@ We'll see there some examples.

-The httrack commandline tool allows (since the 3.30 release) to plug external functions to various callbacks defined in httrack.
-See also: the httrack-library.h prototype file, and the callbacks-example.c given in the httrack archive.
+The httrack commandline tool allows (since the 3.30 release) to plug external functions to various callbacks defined in httrack. +The 3.41 release introduces a cleaned up verion of callbacks, with two major changes: +
+References: + + +
+Below the list of functions to be defined in the module (plugin).
+
+ + + + + + + + + +
module function namefunction descriptionfunction signature
hts_plug +The module entry point. The opt structure can be used to plug callbacks, using the CHAIN_FUNCTION() macro helper. The argv optional argument is the one passed in the commandline as --wrapper parameter.
return value: 1 upon success, 0 upon error (the mirror will then be aborted)
+ +
+Wrappers can be plugged inside hts_plug() using:
+ +CHAIN_FUNCTION(opt, <callback name>, <our callback function name>, <our callback function optional custom pointer argument>); + +
+ +
Example: +
-httrack --wrapper check-html=callback:process_file .. +CHAIN_FUNCTION(opt, check_html, process, userdef); -
-With the callback.so (or callback.dll) module defined as below: +
-
-int process_file(char* html, int len, char* url_adresse, char* url_fichier) {
-  printf("now parsing %s%s..\n", url_adresse, url_fichier);
-  strcpy(currentURLBeingParsed, url_adresse);
-  strcat(currentURLBeingParsed, url_fichier);
-  return 1;  /* success */
-}
-
+		extern int hts_plug(httrackp *opt, const char* argv);
hts_unplug +The module exit point. To free allocated resources without using global variables, use the uninit callback (see below)extern int hts_unplug(httrackp *opt);
-Below the list of callbacks, and associated external wrappers:
+ +
+Note that all callbacks (except init and uninit) take as first two argument: + + +
+Below the list of callbacks, and associated external wrappers. - - - - - - - - - - - - - - - - - - - - -typedef void (* t_hts_htmlcheck_filesave2)(); - - - - - - - + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
"callback name"callback descriptioncallback function signature
"init"Note: deprecated, should not be used anymore (unsafe callback) - see "start" callback or wrapper_init() module function below this table.Called during initialization ; use of htswrap_add (see httrack-library.h) is permitted inside this function to setup other callbacks.
return value: none		void (* myfunction)(void);
"free"Note: deprecated, should not be used anymore (unsafe callback) - see "end" callback or wrapper_exit() module function below this table.
Called during un-initialization
return value: none		void (* myfunction)(void);
"start"Called when the mirror starts. The opt structure passed lists all options defined for this mirror. You may modify the opt structure to fit your needs. Besides, use of htswrap_add (see httrack-library.h) is permitted inside this function to setup other callbacks.
return value: 1 upon success, 0 upon error (the mirror will then be aborted)		int (* myfunction)(httrackp* opt);
"end"Called when the mirror ends
return value: 1 upon success, 0 upon error (the mirror will then be considered aborted)		int (* myfunction)(void);
"change-options"Called when options are to be changed. The opt structure passed lists all options, updated to take account of recent changes
return value: 1 upon success, 0 upon error (the mirror will then be aborted)		int (* myfunction)(httrackp* opt);
"check-html"Called when a document (which may not be an html document) is to be parsed. The html address points to the document data, of lenth len. The url_adresse and url_fichier are the address and URI of the file being processed
return value: 1 if the parsing can be processed, 0 if the file must be skipped without being parsed		int (* myfunction)(char* html,int len,char* url_adresse,char* url_fichier);
"preprocess-html"Called when a document (which is an html document) is to be parsed (original, not yet modified document). The html address points to the document data address (char**), and the length address points to the lenth of this document. Both pointer values (address and size) can be modified to change the document. It is up to the callback function to reallocate the given pointer (using standard C library realloc()/free() functions), which will be free()'ed by the engine. Hence, return of static buffers is strictly forbidden, and the use of strdup() in such cases is advised. The url_adresse and url_fichier are the address and URI of the file being processed
return value: 1 if the new pointers can be applied (default value)		int (* myfunction)(char** html,int* len,char* url_adresse,char* url_fichier);
"postprocess-html"Called when a document (which is an html document) is parsed and transformed (links rewritten). The html address points to the document data address (char**), and the length address points to the lenth of this document. Both pointer values (address and size) can be modified to change the document. It is up to the callback function to reallocate the given pointer (using standard C library realloc()/free() functions), which will be free()'ed by the engine. Hence, return of static buffers is strictly forbidden, and the use of strdup() in such cases is advised. The url_adresse and url_fichier are the address and URI of the file being processed
return value: 1 if the new pointers can be applied (default value)		int (* myfunction)(char** html,int* len,char* url_adresse,char* url_fichier);
"query"Called when the wizard needs to ask a question. The question string contains the question for the (human) user
return value: the string answer ("" for default reply)		char* (* myfunction)(char* question);
"query2"Called when the wizard needs to ask a questionchar* (* myfunction)(char* question);
"query3"Called when the wizard needs to ask a questionchar* (* myfunction)(char* question);
"loop"Called periodically (informational, to display statistics)
return value: 1 if the mirror can continue, 0 if the mirror must be aborted		int (* myfunction)(lien_back* back,int back_max,int back_index,int lien_tot,int lien_ntot,int stat_time,hts_stat_struct* stats);
"check-link"Called when a link has to be tested. The adr and fil are the address and URI of the link being tested. The passed status value has the following meaning: 0 if the link is to be accepted by default, 1 if the link is to be refused by default, and -1 if no decision has yet been taken by the engine
return value: same meaning as the passed status value ; you may generally return -1 to let the engine take the decision by itself		int (* myfunction)(char* adr,char* fil,int status);
"check-mime"Called when a link download has begun, and needs to be tested against its MIME type. The adr and fil are the address and URI of the link being tested, and the mime string contains the link type being processed. The passed status value has the following meaning: 0 if the link is to be accepted by default, 1 if the link is to be refused by default, and -1 if no decision has yet been taken by the engine
return value: same meaning as the passed status value ; you may generally return -1 to let the engine take the decision by itself		int (* myfunction)(char* adr,char* fil,char* mime,int status);
"pause"Called when the engine must pause. When the lockfile passed is deleted, the function can return
return value: none		void (* myfunction)(char* lockfile);
"save-file"Called when a file is to be saved on disk
return value: none		void (* myfunction)(char* file);
"save-file2"Called when a file is to be saved or checked on disk
The hostname, filename and local filename are given. Two additional flags tells if the file is new (is_new) and is the file is to be modified (is_modified).
(!is_new && !is_modified): the file is up-to-date, and will not be modified
(is_new && is_modified): a new file will be written (or an updated file is being written)
(!is_new && is_modified): a file is being updated (append)
(is_new && !is_modified): an empty file will be written ("do not recatch locally erased files")
return value: none		void (* myfunction)(char* hostname,char* filename,char* localfile,int is_new,int is_modified);
"link-detected"Called when a link has been detected
return value: 1 if the link can be analyzed, 0 if the link must not even be considered		int (* myfunction)(char* link);
"transfer-status"Called when a file has been processed (downloaded, updated, or error)
return value: must return 1		int (* myfunction)(lien_back* back);
"save-name"Called when a local filename has to be processed. The adr_complete and fil_complete are the address and URI of the file being saved ; the referer_adr and referer_fil are the address and URI of the referer link. The save string contains the local filename being used. You may modifiy the save string to fit your needs, up to 1024 bytes (note: filename collisions, if any, will be handled by the engine by renaming the file into file-2.ext, file-3.ext ..).
return value: must return 1		int (* myfunction)(char* adr_complete,char* fil_complete,char* referer_adr,char* referer_fil,char* save);
"send-header"Called when HTTP headers are to be sent to the remote server. The buff buffer contains text headers, adr and fil the URL, and referer_adr and referer_fil the referer URL. The outgoing structure contains all information related to the current slot.
return value: 1 if the mirror can continue, 0 if the mirror must be aborted		int (* myfunction)(char* buff, char* adr, char* fil, char* referer_adr, char* referer_fil, htsblk* outgoing);
"receive-header"Called when HTTP headers are recevived from the remote server. The buff buffer contains text headers, adr and fil the URL, and referer_adr and referer_fil the referer URL. The incoming structure contains all information related to the current slot.
return value: 1 if the mirror can continue, 0 if the mirror must be aborted		int (* myfunction)(char* buff, char* adr, char* fil, char* referer_adr, char* referer_fil, htsblk* incoming);
callback namecallback descriptioncallback function signature
initNote: the use the "start" callback is advised. Called during initialization.
return value: none		void mycallback(t_hts_callbackarg *carg);
uninitNote: the use os the "end" callback is advised.
Called during un-initialization
return value: none		void mycallback(t_hts_callbackarg *carg);
startCalled when the mirror starts. The opt structure passed lists all options defined for this mirror. You may modify the opt structure to fit your needs.
return value: 1 upon success, 0 upon error (the mirror will then be aborted)		int mycallback(t_hts_callbackarg *carg, httrackp* opt);
endCalled when the mirror ends
return value: 1 upon success, 0 upon error (the mirror will then be considered aborted)		int mycallback(t_hts_callbackarg *carg, httrackp* opt);
choptCalled when options are to be changed. The opt structure passed lists all options, updated to take account of recent changes
return value: 1 upon success, 0 upon error (the mirror will then be aborted)		int mycallback(t_hts_callbackarg *carg, httrackp* opt);
preprocessCalled when a document (which is an html document) is to be parsed (original, not yet modified document). The html address points to the document data address (char**), and the length address points to the lenth of this document. Both pointer values (address and size) can be modified to change the document. It is up to the callback function to reallocate the given pointer (using the hts_realloc()/hts_free() library functions), which will be free()'ed by the engine. Hence, return of static buffers is strictly forbidden, and the use of hts_strdup() in such cases is advised. The url_address and url_file are the address and URI of the file being processed
return value: 1 if the new pointers can be applied (default value)		int mycallback(t_hts_callbackarg *carg, httrackp* opt, char** html, int* len, const char* url_address, const char* url_file);
postprocessCalled when a document (which is an html document) is parsed and transformed (links rewritten). The html address points to the document data address (char**), and the length address points to the lenth of this document. Both pointer values (address and size) can be modified to change the document. It is up to the callback function to reallocate the given pointer (using the hts_realloc()/hts_free() library functions), which will be free()'ed by the engine. Hence, return of static buffers is strictly forbidden, and the use of hts_strdup() in such cases is advised. The url_address and url_file are the address and URI of the file being processed
return value: 1 if the new pointers can be applied (default value)		int mycallback(t_hts_callbackarg *carg, httrackp* opt, char** html, int* len, const char* url_address, const char* url_file);
check_htmlCalled when a document (which may not be an html document) is to be parsed. The html address points to the document data, of lenth len. The url_address and url_file are the address and URI of the file being processed
return value: 1 if the parsing can be processed, 0 if the file must be skipped without being parsed		int mycallback(t_hts_callbackarg *carg, httrackp* opt, char* html, int len, const char* url_address, const char* url_file);
queryCalled when the wizard needs to ask a question. The question string contains the question for the (human) user
return value: the string answer ("" for default reply)		const char* mycallback(t_hts_callbackarg *carg, httrackp* opt, const char* question);
query2Called when the wizard needs to ask a questionconst char* mycallback(t_hts_callbackarg *carg, httrackp* opt, const char* question);
query3Called when the wizard needs to ask a questionconst char* mycallback(t_hts_callbackarg *carg, httrackp* opt, const char* question);
loopCalled periodically (informational, to display statistics)
return value: 1 if the mirror can continue, 0 if the mirror must be aborted		int mycallback(t_hts_callbackarg *carg, httrackp* opt, lien_back* back, int back_max, int back_index, int lien_tot, int lien_ntot, int stat_time, hts_stat_struct* stats);
check_linkCalled when a link has to be tested. The adr and fil are the address and URI of the link being tested. The passed status value has the following meaning: 0 if the link is to be accepted by default, 1 if the link is to be refused by default, and -1 if no decision has yet been taken by the engine
return value: same meaning as the passed status value ; you may generally return -1 to let the engine take the decision by itself		int mycallback(t_hts_callbackarg *carg, httrackp* opt, const char* adr, const char* fil, int status);
check_mimeCalled when a link download has begun, and needs to be tested against its MIME type. The adr and fil are the address and URI of the link being tested, and the mime string contains the link type being processed. The passed status value has the following meaning: 0 if the link is to be accepted by default, 1 if the link is to be refused by default, and -1 if no decision has yet been taken by the engine
return value: same meaning as the passed status value ; you may generally return -1 to let the engine take the decision by itself		int mycallback(t_hts_callbackarg *carg, httrackp* opt, const char* adr, const char* fil, const char* mime, int status);
pauseCalled when the engine must pause. When the lockfile passed is deleted, the function can return
return value: none		void mycallback(t_hts_callbackarg *carg, httrackp* opt, const char* lockfile);
filesaveCalled when a file is to be saved on disk
return value: none		void mycallback(t_hts_callbackarg *carg, httrackp* opt, const char* file);
filesave2Called when a file is to be saved or checked on disk
The hostname, filename and local filename are given. Two additional flags tells if the local file is new (is_new), if the local file is to be modified (is_modified), and if the file was not updated remotely (not_updated).
(!is_new && !is_modified): the file is up-to-date, and will not be modified
(is_new && is_modified): a new file will be written (or an updated file is being written)
(!is_new && is_modified): a file is being updated (append)
(is_new && !is_modified): an empty file will be written ("do not recatch locally erased files")
not_updated: the file was not re-downloaded because it was up-to-date (no data transfered again)

return value: none		void mycallback(t_hts_callbackarg *carg, httrackp* opt, const char* hostname, const char* filename, const char* localfile, int is_new, int is_modified, int not_updated);
linkdetectedCalled when a link has been detected
return value: 1 if the link can be analyzed, 0 if the link must not even be considered		int mycallback(t_hts_callbackarg *carg, httrackp* opt, char* link);
linkdetected2Called when a link has been detected
return value: 1 if the link can be analyzed, 0 if the link must not even be considered		int mycallback(t_hts_callbackarg *carg, httrackp* opt, char* link, const const char* tag_start);
xfrstatusCalled when a file has been processed (downloaded, updated, or error)
return value: must return 1		int mycallback(t_hts_callbackarg *carg, httrackp* opt, lien_back* back);
savenameCalled when a local filename has to be processed. The adr_complete and fil_complete are the address and URI of the file being saved ; the referer_adr and referer_fil are the address and URI of the referer link. The save string contains the local filename being used. You may modifiy the save string to fit your needs, up to 1024 bytes (note: filename collisions, if any, will be handled by the engine by renaming the file into file-2.ext, file-3.ext ..).
return value: must return 1		int mycallback(t_hts_callbackarg *carg, httrackp* opt, const char* adr_complete, const char* fil_complete, const char* referer_adr, const char* referer_fil, char* save);
sendheadCalled when HTTP headers are to be sent to the remote server. The buff buffer contains text headers, adr and fil the URL, and referer_adr and referer_fil the referer URL. The outgoing structure contains all information related to the current slot.
return value: 1 if the mirror can continue, 0 if the mirror must be aborted		int mycallback(t_hts_callbackarg *carg, httrackp* opt, char* buff, const char* adr, const char* fil, const char* referer_adr, const char* referer_fil, htsblk* outgoing);
receiveheadCalled when HTTP headers are recevived from the remote server. The buff buffer contains text headers, adr and fil the URL, and referer_adr and referer_fil the referer URL. The incoming structure contains all information related to the current slot.
return value: 1 if the mirror can continue, 0 if the mirror must be aborted		int mycallback(t_hts_callbackarg *carg, httrackp* opt, char* buff, const char* adr, const char* fil, const char* referer_adr, const char* referer_fil, htsblk* incoming);
detectCalled when an unknown document is to be parsed. The str structure contains all information related to the document.
return value: 1 if the type is known and can be parsed, 0 if the document type is unknown		int mycallback(t_hts_callbackarg *carg, httrackp* opt, htsmoduleStruct* str);
parseThe str structure contains all information related to the document.
return value: 1 if the document was successfully parsed, 0 if an error occured		int mycallback(t_hts_callbackarg *carg, httrackp* opt, htsmoduleStruct* str);


-Below additional function names that can be defined inside the module (DLL/.so):
- - - +Note: the optional libhttrack-plugin module (libhttrack-plugin.dll or libhttrack-plugin.so), if found in the library environment, is loaded automatically, and its hts_plug() function being called.
- - - -
"module function name"function description
int function-name_init(char *args);Called when a function named function-name is extracted from the current module (same as wrapper_init). The optional args provides additional commandline parameters. Returns 1 upon success, 0 if the function should not be extracted.
int wrapper_init(char *fname, char *args);Called when a function named fname is extracted from the current module. The optional args provides additional commandline parameters. Besides, use of htswrap_add (see httrack-library.h) is permitted inside this function to setup other callbacks. Returns 1 upon success, 0 if the function should not be extracted.
int wrapper_exit(void);Called when the module is unloaded. The function should return 1 (but the result is ignored).
- -

-Below additional function names that can be defined inside the optional libhttrack-plugin module (libhttrack-plugin.dll or libhttrack-plugin.so) searched inside common library path:
- - - +
+An example is generally more efficient than anything else, so let's write our first module, aimed to stupidely print all parsed html files: +
"module function name"function description
+
+
+/* system includes */
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+
+/* standard httrack module includes */
+#include "httrack-library.h"
+#include "htsopt.h"
+#include "htsdefines.h"
+
+/* local function called as "check_html" callback */
+static int process_file(t_hts_callbackarg /*the carg structure, holding various information*/*carg, /*the option settings*/httrackp *opt, 
+                        /*other parameters are callback-specific*/
+                        char* html, int len, const char* url_address, const char* url_file) {
+  void *ourDummyArg = (void*) CALLBACKARG_USERDEF(carg);    /*optional user-defined arg*/
+
+  /* call parent functions if multiple callbacks are chained. you can skip this part, if you don't want previous callbacks to be called. */
+  if (CALLBACKARG_PREV_FUN(carg, check_html) != NULL) {
+    if (!CALLBACKARG_PREV_FUN(carg, check_html)(CALLBACKARG_PREV_CARG(carg), opt,
+                                                html, len, url_address, url_file)) {
+        return 0;  /* abort */
+      }
+  }
+
+  printf("file %s%s content: %s\n", url_address, url_file, html);
+  return 1;  /* success */
+}
+
+/* local function called as "end" callback */
+static int end_of_mirror(t_hts_callbackarg /*the carg structure, holding various information*/*carg, /*the option settings*/httrackp *opt) {
+  void *ourDummyArg = (void*) CALLBACKARG_USERDEF(carg);    /*optional user-defined arg*/
+
+  /* processing */
+  fprintf(stderr, "That's all, folks!\n");
+
+  /* call parent functions if multiple callbacks are chained. you can skip this part, if you don't want previous callbacks to be called. */
+  if (CALLBACKARG_PREV_FUN(carg, end) != NULL) {
+    /* status is ok on our side, return other callabck's status */
+    return CALLBACKARG_PREV_FUN(carg, end)(CALLBACKARG_PREV_CARG(carg), opt);
+  }
+
+  return 1;  /* success */
+}
+
+/*
+module entry point
+the function name and prototype MUST match this prototype
+*/
+EXTERNAL_FUNCTION int hts_plug(httrackp *opt, const char* argv) {
+  /* optional argument passed in the commandline we won't be using here */
+  const char *arg = strchr(argv, ',');
+  if (arg != NULL)
+    arg++;
+
+  /* plug callback functions */
+  CHAIN_FUNCTION(opt, check_html, process_file, /*optional user-defined arg*/NULL);
+  CHAIN_FUNCTION(opt, end, end_of_mirror, /*optional user-defined arg*/NULL);
+
+  return 1;  /* success */
+}
+
+/*
+module exit point
+the function name and prototype MUST match this prototype
+*/
+EXTERNAL_FUNCTION int hts_unplug(httrackp *opt) {
+  fprintf(stder, "Module unplugged");
+
+  return 1;  /* success */
+}
+
+
+ +
+Compile this file ; for example: +
+ +gcc -O -g3 -shared -o mylibrary.so myexample.c + +
+and plug the module using the commandline ; for example: +
+ +httrack --wrapper mylibrary http://www.example.com + +
+or, if some parameters are desired: +
+ +httrack --wrapper mylibrary,myparameter-string http://www.example.com + +
+(the "myparameter-string" string will be available in the 'arg' parameter passed to the hts_plug entry point) +
-void plugin_init(void);Called if the module (named libhttrack-plugin.(so|dll)) is found in the library path. Use of htswrap_add (see httrack-library.h) is permitted inside this function to setup other callbacks. - - - -

-

@@ -202,7 +332,7 @@ Below additional function names that can be defined inside the optional libhttra - +
-- cgit v1.2.3