Maybe... Doubtfully... But I wanted to give it a try since I'm tired of my glorious pronouncements garbled in &39
If I actually fixed it then I should get a medal for best coding by someone who can't code. My logic was that WHISPER works just fine with apostrophes and ampersands but shuttle call reasons, announcements, etc. would get that garbled text when using apostrophes and ampersands.
With whispers trim is on the "inside", with stripped_input that is used for these announcements trim is on the "outside". It makes intuitive sense to me that whoever wrote that stripped_input comment might have gotten things flipped since if trim is on the inside, and it allows single symbols to pass through, then there's nothing to filter html_encode from spewing that garbage out.
If this isn't the correct fix then at least I should be close enough that someone can point me in the right direction.
#9902 replaced /proc/strip_html_properly but the new code didn't actually have the character length check that the proc had.
I don't know if trim() is the best place for it or to put it in its own proc but I assume this was Miauw's intent because he is passing max_length into it.
It was not as proper as the name would imply, and experience and amused redditors have taught me that writing a proper HTML-scrubbing function is a very hard task indeed.
* Added stripping everything between http and ://, to prevent links being passed to Byond text window
* Stripped an unsafe inputting for advanced diseases
- Moved the "parsepencode" proc from **newscaster.dm** and **PDA.dm** to
**text.dm** as a global proc.
- Made the "parsepencode" proc accept only a single agrument if desired.
- Removed the font constants from **newscaster.dm**, **PDA.dm**,
**paper.dm**, **paperbin.dm**, and **photocopier.dm**.
- Replaced aforementioned constants with define'ers in **misc.dm**.
- Created a copy of stripped_input called stripped_multiline_input that
does the same, expect it is a multiline input.
- Makes the writing input box multi-line again.
- Renaming paper now uses the stripped_input.
- Fixes https://github.com/tgstation/-tg-station/issues/7132
Regex is a robust text-parsing mini-language. It can perform complex and dynamic searches and return the results in an easy to read format. This implementation also uses Boost's extended formatting syntax which gives you a flexible way to conditionally replace complex strings with only a few lines of code
Source code and license info can be found here: http://code.google.com/p/byond-regex/
Licensed under LGPL.
This uses the perl regex syntax (with some extensions) in non-recursive mode (as recursion has additional overheads).
Information on this syntax can be found here: http://www.boost.org/doc/libs/1_55_0/libs/regex/doc/html/boost_regex/syntax/perl_syntax.html
There are many introductions to regex on the internet. Here is the one I learned from: http://www.regular-expressions.info/tutorial.html
Note: I have added #define USE_BYGEX in code/_compile_options.dm. Simply comment out the define if you encounter any problems.
Due to byond being unable to communicate with a dll through anything but C-strings, the results are returned to byond in a string similar to those you get with list2params. Since byond uses parameters a lot, its params2list() proc is fairly streamlined, so it's pretty much the fastest way you could do this with byond. This data is then stored in a datum/regex object for easy use. See demo.dm datum/regex/proc/report() for an example.
Basically, copytext's first argument is inclusive, but its 2nd argument isn't.
so copytext("123456789", 2, 8) returns 234567.
When I googled "byond copytext" (to see how it worked in byond and if it used 1 or 0 based indexing) the first link is http://www.byond.com/docs/guide/chap15.html and this does nothing to spell this out.
The reference guide (that i should have used) spells this out more clearly, but it is confusing none the less: http://www.byond.com/docs/ref/info.html#/proc/copytext
Changed all the hair/facial_hair/eye sprites to use a white base.
Changed the preview screen, in the preference menu, to use ICON_MULTIPLY to be able to use the new white sprites.
Removed the unused and broken cloaking item.
Ninjas now use the alpha variable, removed most of the checks for it.
Updated the documentation a bit.
Added some helper procs for later.
Added helper: /proc/findchar(haystack, needles, start=1, end=0)
works like findtext except it finds the first occurrence of one of the characters from the needles string, within haystack.
Permissions can now be removed as well as added, by replacing the + with a -. e.g. +BAN adds bans permissions, -BAN removes them. This applies to admin_ranks.txt and the permissions panel.
Verb overrides are now possible for admins_ranks. Specific verbs can be forced-on, or forced-off via keywords like +/client/proc/mimespeak or -/client/verb/ooc. This applies both to admin_ranks.txt and the permissions panel. SQL system is not compatible.
admin_ranks are now datums. This means admin_ranks actually behave more like permission groups as intended.
When you temporarily modify an admin's permissions via the permissions panel, their rank_name is appended with "([ckey])". This prevents modifications to their rights affecting everyone.
admin rank names now support - _ and @ characters (since ckeyEx() does not strip these).
SQL permissions system was modified. SQL databases will have to be updated or they will no longer work.
WARNING: admin_ranks.txt and admins.txt format has changed slightly! You will need to redo those txt files!
Fix for Overpowered mutations (HULK and TELE) not manifesting. A typo was the cause.
Admin-spawnable injectors always succeed. This code already existed (I believe this is how they used to function), but it did not work as the argument was not passed correctly. This also removed snowflake code for the darkbundle telemut injector.
Thanks to Giacom, wouldn't have found these without him.
DNA modification uses normally distributed random numbers. Radiation strength dictates the standard deviation of the change in the hex character which is hit (higher output means greater chance for a large change) . Similarly, radiation duration dictates how likely we are to hit the hex-character we clicked on (longer duration means more likely to hit). Irradiation is strength*duration. All balancing is done via multiplier defines so you can rebalance it easily.
DNA blocks and structuring all use defines. Making modification/expansion easier. I'll likely expand this into a datum-based system to allow more interesting features, reduce code further and allow admins to interact with the way dna strings behave.
DNA strings can be spliced together using merge_text().
e.g. string 1: "Hello World"
string 2: "Seeya______"
result: "Seeya World"
This isn't used except for admin-spawnable SE injectors at the moment.
r_hair, g_hair, b_hair, r_facial, g_facial, b_facial, r_eyes, g_eyes, b_eyes were removed and made into 3 short hex-colors.
Skin tones now support colours other than shades of brown. I've had to restrict it heavily until other stuff is done. Skin tones include Albino Caucasian, Oriental, Mediterranean, etc.
Data disks and DNA injectors were reworked to use associative lists so transferring data is just a matter of doing list.Copy()
var/dna is now defined at /mob/living/carbon level. Only monkeys and humans may have dna currently. Support is there for all carbon-based lifeforms to have dna.
DNA modifier console has almost all controls on one screen.
UIs and UEs can be injected separately (appearance and name, respectively)
dna helper procs like ready_dna() and such were changed to make them more versatile. There is now a hardset_dna() proc as an alternative to ready_dna which can initialize dna with properties passed into it or update an existing dna string (useful for cloning and antag spawning)
Every block of SEs are in randomised positions.
Disabled automatic logging of world.log, as it produced undesirable behaviour.
Mr Muggles and God Emperor of Mankind disks removed.
Floor() removed. (it was completely uneccesary, that is what round() is).
Fixed spelling mistakes in modularchangling.dm (thanks tenebrosity)
Tanning removed from beaches (again)
Experimental: monkeys and humans do not have dna until first attempt to read dna (using check_dna_integrity(mob))
This is mainly due to the way everything is hardcoded into New().
Changelog.html updated
Signed-off-by: carnie <elly1989@rocketmail.com>
>dd_replacetext and dd_replacetext_case are now replacetext and replacetextEx respectively.
>replacetext and replacetextEx works using strings only; it doesn't convert it into a list and then back into a string.
Fixes some dumb-dumbs in textlist and text2listEx
> "<" where there should be a "<="
> no else case for when the separator is longer than the text (causing empty lists to be returned)
git-svn-id: http://tgstation13.googlecode.com/svn/trunk@4971 316c924e-a436-60f5-8080-3fe189b3f50e
>Replaced dd_text2list, dd_text2listcase, tg_text2listcase and tg_text2list with text2list and text2listEx. text2list will return a list of each and every character in the string if you set separator=""
>added return_file_text(filepath) which returns text from a file after doing some checks: does the file exist? is the file empty? It prints helpful error messages to the world.log if it runs into problems
>Replaced dd_file2list(filepath, seperator) with file2list(filepath, seperator). It just calls text2list(return_file_text(filepath), seperator). rather than copypasta
>Replaced time_stamp() so it's not as retarded
>Lots of the world setup stuff uses file2list now, rather than file2text -> sanity -> text2list
>Added error() warning() testing() procs. These print messages to world.log with a prefix. e.g. ## ERROR: msg.
git-svn-id: http://tgstation13.googlecode.com/svn/trunk@4948 316c924e-a436-60f5-8080-3fe189b3f50e
AnturK
Bjorn Neergaard
Firecage
Robustin
Cheridan
Miauw
Menshin
Menshin
Deantwo
hornygranny
Razharas
carnie
MrStonedOne
Giacomand
YotaXP
giacomand@gmail.com
elly1989@rocketmail.com