split out parser modes into their own files

This moves all the parser classes into their own namespace and files.
Next up are the handler classes.

I'm not sure about the namespace, yet. A nested na

split out parser modes into their own files

This moves all the parser classes into their own namespace and files.
Next up are the handler classes.

I'm not sure about the namespace, yet. A nested namepspace Parser\Modes
would probably make more sense... we'll see.

This also removes the duplicated coded in the Plugin mode. We now use
the plugin trait and can inherit from AbstractMode instead.

show more ...

visibility definition for the renderers

I made a lot of things public that probaly should be protected. But many
syntax plugins do access renderer mechanisms directly, so better stay on
the safe sid

visibility definition for the renderers

I made a lot of things public that probaly should be protected. But many
syntax plugins do access renderer mechanisms directly, so better stay on
the safe side here.

The base renderer is now abstract.

show more ...

fixed broken tests

I also introduced an auto loaded namespace for the tests.

visibility fixes

First start at declaring visibilites for methods and properties. Still
missing: the parser/renderer stuff and the plugins

feat: improve search as a primary means to create new pages

Specifically, this pull request checks if the current query is already a
valid pageid and then return that as suggested pageid.

Also, tes

feat: improve search as a primary means to create new pages

Specifically, this pull request checks if the current query is already a
valid pageid and then return that as suggested pageid.

Also, tests are added for this method.

Fixes #2355

PS: We may want to somehow better educate/nudge users to use the better
way of creating pages by creating links on existing pages.

show more ...

Prevent HTML and JS injection in section names

Before this change, HTML and some JS code (as far as it was not escaped
by json_encode) could be injected into the output as the closing pattern
that i

Prevent HTML and JS injection in section names

Before this change, HTML and some JS code (as far as it was not escaped
by json_encode) could be injected into the output as the closing pattern
that is checked by the regex is not escaped in JSON (see test case).

show more ...

relax geshi options a bit more

Quotes are now optional, the options are spearated by space

renamed ActionRouter::checkPermissions to checkPreconditions

As discussed in #1933

feat: add a simple unparser for parsed search queries

To allow creating links with manipulated versions of the current search
query an unparser is necessary. However, the current output of ft_queryP

feat: add a simple unparser for parsed search queries

To allow creating links with manipulated versions of the current search
query an unparser is necessary. However, the current output of ft_queryParser
makes some advanced features hard to detect. Therefore the new
ft_queryUnparser_simple cannot handle negated phrases and `OR` searches.

It should still cover 98% of search queries.

show more ...

use expectException instead of setExpectedException for PHPUnit

Refactored section edit

The data for the section edit button is now passed as an assoziative array which is
encoded in the '#<!-- EDIT(.*) -->#' placeholder as an JSON array.

add tests to DifferenceEngine and optimize each removal in it

fix: prevent two selected options, b/c apparently 'String' == 0 ��

There was a bug, where the options array ['Auto', 0, 1] would result in
HTML option tags where both the 'Auto' and the 0 option we

fix: prevent two selected options, b/c apparently 'String' == 0 ��

There was a bug, where the options array ['Auto', 0, 1] would result in
HTML option tags where both the 'Auto' and the 0 option were selected.

show more ...

fix file name
closes #2129

proper handling of "0" as a search query in media manager + test case for this

create some unit tests that test searching in media manager

test: add tests for the section edit regex

another go at working around the networink troubles at travis #2109

show more info on failed HTTP requests

hopefully this will help with figuring out why our HTTP tests sometimes
fail on travis.

some cleanup fpr set_metadata test

added one more test for internal links

just to be sure, we're not breaking something simple here.

parsertests: replaced var keywords and added type hints

Add tests for array_replace part of set_metadata

response on review

cleaning