* Move configuration in its own class in order to reduce the verbosity of index.php
* Add authentication mechanism using HTTP auth
* Add a method to get the config parameters
* Remove the installation checks from the index page
* Log all failed authentication attempts
cURL is a powerful library specifically designed to connect to many
different types of servers with different types of protocols. For
more detailed information refer to the PHP cURL manual:
- http://php.net/manual/en/book.curl.php
Due to this change some parameters for the getContents function were
necessary (also applies to getSimpleHTMLDOM and getSimpleHTMLDOMCached):
> $use_include_path removed
This parameter has never been used and doesn't even make sense in
this context; If set to true file_get_contents would also search
for files in the include_path (specified in php.ini).
> $context replaced by $header and $opts
The $context parameter allowed for customization of the request in
order to change how file_get_contents would acquire the data (i.e.
using POST instead of GET, sending custom header, etc...)
cURL also provides facilities to specify custom headers and change
how it communicates to severs. cURL, however, is much more advanced.
- $header is an optional parameter (empty by default). It receives
an array of strings to send in the HTTP request header.
See 'CURLOPT_HTTPHEADER':
"An array of HTTP header fields to set, in the format
array('Content-type: text/plain', 'Content-length: 100')"
- php.net/manual/en/function.curl-setopt.php
- $opts is an optional parameter (empty by default). It receives
an array of options, where each option is a key-value-pair of
a cURL option (CURLOPT_*) and it's associated parameter. This
parameter accepts any of the CURLOPT_* settings.
Example (sending POST instead of GET):
$opts = array(
CURLOPT_POST => 1,
CURLOPT_POSTFIELDS => '&action=none'
);
$html = getContents($url, array(), $opts);
Refer to the cURL setopt manual for more information:
- php.net/manual/en/function.curl-setopt.php
> $offset and $maxlen removed
These options were supported by file_get_contents, but there doesn't
seem to be an equivalent in cURL. Since no caller uses them they are
safe to remove.
Compressed data / Encoding
By using cURL instead of file_get_contents RSS-Bridge no longer has
to handle compressed data manually.
See 'CURLOPT_ENCODING':
"[...] Supported encodings are "identity", "deflate", and "gzip".
If an empty string, "", is set, a header containing all supported
encoding types is sent."
- http://php.net/manual/en/function.curl-setopt.php
Notice: By default all encoding types are accepted (""). This can
be changed by setting a custom option via $opts.
Example:
$opts = array(CURLOPT_ENCODING => 'gzip');
$html = getContents($url, array(), $opts);
Proxy
The proxy implementation should still work, but there doesn't seem
to be an equivalent for 'request_fulluri = true'. To my understanding
this isn't an issue because cURL knows how to handle proxy communication.
* [BridgeAbstract] Implement customizable cache timeout
The customizable cache timeout is used instead of the default cache
timeout (CACHE_TIMEOUT) if specified by the caller.
* [index] Add new global parameter '_cache_timeout'
The _cache_timeout parameter is an optional parameter that can be
used to specify a custom cache timeout. This option is enabled by
default.
It can be disabled using the named constant 'CUSTOM_CACHE_TIMEOUT'
which supports two states:
> true: Enabled (default)
> false: Disabled
* [BridgeAbstract] Change scope of 'getCacheTimeout' to public
* [html] Add cache timeout parameter to all bridges
The timeout parameter only shows if CUSTOM_CACHE_TIMEOUT has been set
to true. The default value is automatically set to the value specified
in the bridge.
* [index] Disable custom cache timeout by default
Providing no value for an optional numeric parameter results in
error "Parameter *** is invalid!"
This is caused by the validation function ignoring the 'required'
attribute when loading and checking input parameters.
This commit adds checks to determine whether the 'required' attri-
bute is defined and active before returning the error message.
References #570:
Whitespace at the beginning of feeds causes parsing errors. This is
an example using an ill-formatted RSS feed:
"XML or text declaration not at start of entity"
-- https://validator.w3.org
This commit automatically removes all proceeding and trailing white-
space from the source content before resume parsing.
Matching whitelisted bridges using a case-insensitive match makes
sense for following reasons:
- Wrong upper/lower case spelling in the whitelist is not easily
discovered. Example: Misspelling 'Youtube' as 'YouTube' will not
show the 'Youtube' bridge (while it is expected to show)
- Two bridges with the same name but different letter casing are
discouraged to prevent confusion and keep the project compatible
with Windows machines
- Do not add spaces after opening or before closing parenthesis
// Wrong
if( !is_null($var) ) {
...
}
// Right
if(!is_null($var)) {
...
}
- Add space after closing parenthesis
// Wrong
if(true){
...
}
// Right
if(true) {
...
}
- Add body into new line
- Close body in new line
// Wrong
if(true) { ... }
// Right
if(true) {
...
}
Notice: Spaces after keywords are not detected:
// Wrong (not detected)
// -> space after 'if' and missing space after 'else'
if (true) {
...
} else{
...
}
// Right
if(true) {
...
} else {
...
}
Add additional information to error message:
- Name of the bridge
- Possible solutions
- Error description
- Error code
- Error message
* Output type changed from 'text' to 'html'
* Added styles for the error page
* Added a button to remotely open a GitHub issue
Closes#525
Re-requesting the same feed in normal mode (not debug mode)
returns the cached version. The name and URI of the feed
however was not returned.
This adds checks to getName() and getURI() in order to return
the cached infos when using the cached version.
Notice: For this to work correctly all bridges must call to
parent::getName() and parent::getURI() respectively if the
queriedContext is not defined. Example:
switch($this->queriedContext){
case 'My context':
// do your stuff here!
break;
default: parent::getName();
}
* add function getExtraInfos() to BridgeAbstract
* replace call to $bridge->getName() and $bridge->getURI() by $bridge->getExtraInfos()
replace call to $bridge->getName() and $bridge->getURI() by $bridge->getExtraInfos() defined by default in BridgeAbstract.
So we could pass additionals ExtraInfos from custom bridges to custom formats.
Previously BridgeAbstract needed to know which exact implementation
of CacheInterface was used (since we only got one right now its
not a problem). Initializing the cache in index.php instead allows
to change cache types more easily.
The function 'prepare' previously implemented in CacheAbstract
is specifically required for FileCache and thus belongs to FileCache.
Since this change removes all code from CacheAbstract, it can be
removed completely.
