1 2012-06-04T00:33:15 <breton> http://dpaste.com/754935/
2 2012-06-04T00:36:16 <breton> that's an old wiki item from Apr 3, 2012 7:40:24 PM
3 2012-06-04T00:39:43 <breton> hmm, it happens when I do {{test}}, where "test" is http://dpaste.com/754937/
4 2012-06-04T00:42:12 <breton> from 1 to 15 is meta and from 17 to 22 is item itself
5 2012-06-04T02:40:56 *** dwcramer
6 2012-06-04T03:42:14 *** jaiditya
7 2012-06-04T04:10:51 *** jaiditya|2
8 2012-06-04T04:48:15 *** dwcramer
9 2012-06-04T05:01:15 *** dwcramer
10 2012-06-04T05:27:27 *** dwcramer
11 2012-06-04T09:54:20 <ThomasWaldmann> moin
12 2012-06-04T10:00:21 <ThomasWaldmann> breton: can you reinstall emeraldtree and try again?
13 2012-06-04T10:01:41 <ThomasWaldmann> (or rather uninstall it, and then pip install --download-cache=dlc -e . )
14 2012-06-04T10:08:11 <ThomasWaldmann> (you should have emeraldtree 0.9.1 afterwards)
15 2012-06-04T10:12:13 <ThomasWaldmann> btw, all students are advised to do that (after updating from main repo)
16 2012-06-04T11:21:06 *** greg_f
17 2012-06-04T13:11:35 <MaikuMori> moin
18 2012-06-04T13:18:44 <ThomasWaldmann> hi MaikuMori
19 2012-06-04T13:19:03 <MaikuMori> hey
20 2012-06-04T13:20:28 <MaikuMori> Sorry that I missed the meeting
21 2012-06-04T13:24:32 <ThomasWaldmann> did you agre on the meeting today with eSyr-ng ?
22 2012-06-04T13:24:36 <ThomasWaldmann> +e
23 2012-06-04T13:25:05 <MaikuMori> nope
24 2012-06-04T13:27:09 * eSyr-ng here
25 2012-06-04T13:27:37 <ThomasWaldmann> well, if you both have time...
26 2012-06-04T13:27:50 <MaikuMori> I do
27 2012-06-04T13:27:57 * eSyr-ng too
28 2012-06-04T13:30:55 * ThomasWaldmann is currently waiting to get his car maintenance finished, so /me too :)
29 2012-06-04T13:32:57 <MaikuMori> Well I should have some time now as I have only exams and 1 bigger assignment left
30 2012-06-04T13:33:25 <eSyr-ng> good.
31 2012-06-04T13:33:31 <MaikuMori> Hopefully I'll catch up with the stuff I had to do up till now.
32 2012-06-04T13:34:22 <MaikuMori> I've started making exmaple plugin
33 2012-06-04T13:34:24 * eSyr-ng is looking at ep for open questions
34 2012-06-04T13:34:58 <MaikuMori> Going to be just barebone pyhton package and then I'm thinking about looking into loading.
35 2012-06-04T13:35:25 <MaikuMori> More or less all the sections without conclusions need some more discussion
36 2012-06-04T13:35:49 <ThomasWaldmann> did you think about whether the "docs containing" example plugin makes sense?
37 2012-06-04T13:36:12 <MaikuMori> I believe having a wikipage would be better
38 2012-06-04T13:36:47 <MaikuMori> + if I were to try to stuff everything in one plugin it would becone quite bloated in my opinion
39 2012-06-04T13:37:21 <MaikuMori> Though I do think there should at least one example plugin, just maybe not with all functionality
40 2012-06-04T13:37:24 <ThomasWaldmann> no wiki page
41 2012-06-04T13:37:58 * eSyr-ng currently missing purpose of wiki page
42 2012-06-04T13:38:00 <ThomasWaldmann> we had that in moin 1.9, but in moin2 we have sphinx docs
43 2012-06-04T13:38:06 <eSyr-ng> wikipage as documentation for plugin?
44 2012-06-04T13:38:14 <MaikuMori> Could document the API there then
45 2012-06-04T13:38:21 <MaikuMori> as docs for api
46 2012-06-04T13:38:25 <ThomasWaldmann> yes, plugin api docs, examples
47 2012-06-04T13:38:40 <eSyr-ng> well, i think sphinx would be fine.
48 2012-06-04T13:38:44 <ThomasWaldmann> the docs then can refer directly to the code parts of the same plugin
49 2012-06-04T13:39:06 <eSyr-ng> wiki integration is good idea, but it can be made as part of integration of sphinx docs into wiki (if would ever be)
50 2012-06-04T13:39:25 <MaikuMori> It's just that I think it should be a section in moin docs.
51 2012-06-04T13:39:41 <ThomasWaldmann> eSyr-ng: well, that is not in scope of THIS project
52 2012-06-04T13:39:55 <MaikuMori> Because that's where I would look for plugin API if I wanted to develop a plugin.
53 2012-06-04T13:40:08 <eSyr-ng> yes, that's what i'm trying to pointing out :)
54 2012-06-04T13:41:04 <eSyr-ng> MaikuMori: so, don't bother yourself some separate documentation effort, you can put everything into code.
55 2012-06-04T13:41:17 <ThomasWaldmann> MaikuMori: then we have to bundle the example plugin into moin
56 2012-06-04T13:42:19 <MaikuMori> Well the API part will be in moin package. I could just document it there.
57 2012-06-04T13:42:33 <MaikuMori> And then leave the example plugin with maybe very simple docs
58 2012-06-04T13:42:47 <ThomasWaldmann> ok, whathever works better
59 2012-06-04T13:45:43 <MaikuMori> Any comments on my plan? Make barebone package which does nothing for now, then work on discovery/loading/registration. And at the same time we can work on ep with the other sections.
60 2012-06-04T13:46:26 <MaikuMori> on the*
61 2012-06-04T13:47:03 <eSyr-ng> well, example plugin is a good start to play with.
62 2012-06-04T13:48:04 <MaikuMori> It has to be developed at the same time as main api though
63 2012-06-04T13:48:26 <MaikuMori> apart from barebone package to play with loading/discovery
64 2012-06-04T13:53:30 <ThomasWaldmann> btw, if you work with sphinx, try to adapt to the style of existing docs (e.g. headlines markup)
65 2012-06-04T13:54:02 <MaikuMori> Alright, will do.
66 2012-06-04T13:54:48 <MaikuMori> Should we go trough trivial stuff in codereview (creation of basic package)
67 2012-06-04T13:57:02 <ThomasWaldmann> i think you could just steal some stuff from the moin2 package
68 2012-06-04T13:57:48 <eSyr-ng> :3
69 2012-06-04T13:57:56 <ThomasWaldmann> gtg / brb
70 2012-06-04T13:59:58 <MaikuMori> ThomasWaldmann: already done that :)
71 2012-06-04T14:05:27 *** dwcramer
72 2012-06-04T14:42:53 *** dave_largo
73 2012-06-04T14:58:58 *** DasIch_
74 2012-06-04T15:00:04 *** DasIch
75 2012-06-04T15:00:04 *** DasIch_
76 2012-06-04T15:11:29 *** vpv
77 2012-06-04T15:13:10 *** vpv
78 2012-06-04T15:34:01 *** dwcramer
79 2012-06-04T15:36:30 *** jaiditya|2
80 2012-06-04T15:40:02 *** jaiditya|2
81 2012-06-04T15:40:59 *** kanha_
82 2012-06-04T15:41:34 *** kanha
83 2012-06-04T15:42:54 *** jaiditya
84 2012-06-04T15:45:23 *** jaiditya|2
85 2012-06-04T15:53:27 <jaiditya> moin
86 2012-06-04T16:02:24 <jaiditya> waldi: meeting ?
87 2012-06-04T16:06:54 *** dwcramer
88 2012-06-04T16:17:28 <waldi> yes
89 2012-06-04T16:20:09 <jaiditya> have you seen the changesets
90 2012-06-04T16:20:27 <waldi> i have not seen anything in the public repository
91 2012-06-04T16:21:23 <jaiditya> It was told that the changesets have to be first reviewed by the mentors
92 2012-06-04T16:21:43 <jaiditya> so I was waiting for it
93 2012-06-04T16:28:21 <jaiditya> I am done with the color palette of foobar theme and moved on to further styling.
94 2012-06-04T16:32:48 <jaiditya> waldi: While styling the usersettings I think of converting the tabs into the local panel . What do you think ?
95 2012-06-04T16:34:27 <jaiditya> for doing so we'll have to add a new template to foobar theme for usersettings
96 2012-06-04T16:35:47 <jaiditya> which will extend local.html
97 2012-06-04T16:54:55 <waldi> no, we don't want to review everything. you can ask if you are uncertain, but it is your project. we will do a thorough review anyway of the stuff that you provide as the outcome of the project, however you have to actually publish that regularly
98 2012-06-04T16:57:13 <jaiditya> waldi: ok. so I should push it without review
99 2012-06-04T16:57:44 <eSyr-ng> jaiditya: most review effort will be done during merge, but some intermediate reviews are also useful.
100 2012-06-04T16:58:29 <eSyr-ng> jaiditya: review of every commit is not necessary.
101 2012-06-04T16:58:55 <jaiditya> ok
102 2012-06-04T17:03:03 *** dwcramer
103 2012-06-04T17:04:08 <jaiditya> waldi: what do you think about moving tabs of usersettings to local panel in foobar theme ?
104 2012-06-04T17:06:44 <waldi> not sure what this would change
105 2012-06-04T17:10:38 <jaiditya> I don't think changing layout should affect anything. But if you are not sure I'll test it first
106 2012-06-04T17:13:45 <jaiditya> waldi: one last thing as said by Thomas Waldmann the changesets are very small. So should I combine few sets before pushing ?
107 2012-06-04T17:18:34 *** dwcramer
108 2012-06-04T17:31:26 *** Oglarun
109 2012-06-04T17:34:24 *** dwcramer
110 2012-06-04T17:48:01 *** fdik
111 2012-06-04T17:48:07 <fdik> ThomasWaldmann: Huhu! ;-)
112 2012-06-04T17:48:18 <ThomasWaldmann> jaiditya: use some good size and keep them clean (topic wise). if i made a comment about size, it maybe was rather for review (do not request micro-reviews).
113 2012-06-04T17:48:27 <ThomasWaldmann> hi fdik
114 2012-06-04T17:48:51 *** dwcramer
115 2012-06-04T17:49:09 <ThomasWaldmann> waldi: that you did not notice the review requests is no good sign, they are even in the calendar
116 2012-06-04T17:49:52 <fdik> ThomasWaldmann: the basic idea would be to define a bijective relation between subset(html) and $HOWEVER_MOIN_LANGUAGE_IS_NAMED
117 2012-06-04T17:50:08 <fdik> ThomasWaldmann: how is it named BTW? ;-)
118 2012-06-04T17:50:40 <fdik> ThomasWaldmann: then the idea would be to create a stylesheet which grabs the correct subset out of the text and throws away the rest.
119 2012-06-04T17:50:53 <fdik> ThomasWaldmann: is there a definition like that already?
120 2012-06-04T17:51:03 <ThomasWaldmann> we have tests for roundtrip.
121 2012-06-04T17:51:50 <ThomasWaldmann> and there is no formal definition besides that
122 2012-06-04T17:52:13 <ThomasWaldmann> also, we do not use xslt, in case you mean that
123 2012-06-04T17:52:20 <eSyr-ng> fdik: MoinWikiSyntax
124 2012-06-04T17:52:26 <eSyr-ng> something like this
125 2012-06-04T17:52:27 <fdik> eSyr-ng: THX!
126 2012-06-04T17:52:49 <fdik> eSyr-ng: on http://moinmo.in ?
127 2012-06-04T17:53:01 <ThomasWaldmann> but you said you do not want to work on that, so what's the purpose of this discussion? :)
128 2012-06-04T17:53:01 <eSyr-ng> fdik: my username?
129 2012-06-04T17:53:40 <fdik> eSyr-ng: or do you mean http://moinmo.in/HelpOnMoinWikiSyntax ?
130 2012-06-04T17:53:46 <eSyr-ng> fdik: yes.
131 2012-06-04T17:53:48 <fdik> eSyr-ng: sorry, what is with your username?
132 2012-06-04T17:54:04 <eSyr-ng> fdik: i tried to understand your question :)
133 2012-06-04T17:54:55 <eSyr-ng> fdik: and failed first time.
134 2012-06-04T17:55:19 <fdik> eSyr-ng: oh.
135 2012-06-04T17:55:26 <fdik> eSyr-ng: sorry for that.
136 2012-06-04T17:56:21 <fdik> ThomasWaldmann: first: it's just being curious
137 2012-06-04T17:56:29 <fdik> ThomasWaldmann: second: I didn't say that
138 2012-06-04T17:56:33 <eSyr-ng> fdik: ehrr, no apology needed.
139 2012-06-04T17:57:03 <fdik> ThomasWaldmann: I sead, some days would be OK, but I cannot afford weeks or months, because I have no financing then - have to work ;-)
140 2012-06-04T17:57:16 <eSyr-ng> fdik: and moin2 has slightly different syntax, but differences are minor mostly.
141 2012-06-04T17:57:17 <fdik> s/ea/ai/
142 2012-06-04T17:57:35 <fdik> eSyr-ng: aha. Is there documentation for this already?
143 2012-06-04T17:57:44 <eSyr-ng> mmmm, i suppose so.
144 2012-06-04T17:58:29 <eSyr-ng> fdik: http://readthedocs.org/docs/moin-20/en/latest/user/moinwiki.html
145 2012-06-04T17:59:25 <eSyr-ng> fdik: these differences are mostly cornercase-related, so i suppose no big difference between this document and HelpOnMoinWikiSyntax.
146 2012-06-04T17:59:32 <eSyr-ng> at least, i can't remember any.
147 2012-06-04T18:01:22 <fdik> eSyr-ng: THX!
148 2012-06-04T18:02:33 <fdik> eSyr-ng: with which tool are you parsing that?
149 2012-06-04T18:03:46 <fdik> ThomasWaldmann: is there a reason why you have an HTML view on the pages?
150 2012-06-04T18:03:55 <fdik> ThomasWaldmann: I mean, could it be XML?
151 2012-06-04T18:09:25 *** MattMaker
152 2012-06-04T18:11:27 <fdik> ?
153 2012-06-04T18:11:40 <fdik> Why are you guys doing so much parsing work manually?
154 2012-06-04T18:13:40 * fdik cannot see the reason, maybe it's right in front of my face
155 2012-06-04T18:16:44 <ThomasWaldmann> fdik: we use html5 for the output and generate it with emeraldtree (if that answers your question)
156 2012-06-04T18:17:28 <ThomasWaldmann> and we use the polyglot variant, which is well-formed xml also
157 2012-06-04T18:20:15 <fdik> ThomasWaldmann: I mean, you will have an internal format, won't you?
158 2012-06-04T18:20:27 <fdik> ThomasWaldmann: which you're converting to HTML then
159 2012-06-04T18:20:28 *** dave_largo
160 2012-06-04T18:20:28 *** bilal
161 2012-06-04T18:20:30 <fdik> ThomasWaldmann: or not?
162 2012-06-04T18:22:24 <fdik> ThomasWaldmann: http://hg.moinmo.in/moin/2.0/file/87ade74f956a/MoinMoin/converter/moinwiki_in.py
163 2012-06-04T18:22:50 <fdik> ThomasWaldmann: is that the file where you're parsing your own format or am I mistaken?
164 2012-06-04T18:23:10 <ThomasWaldmann> we have a moin-dom tree in memory that is converted to html5 tree that is serialized on output
165 2012-06-04T18:23:32 <fdik> ThomasWaldmann: OK. Do you have a tip for me where I can find the moin-dom tree?
166 2012-06-04T18:23:36 <ThomasWaldmann> that's moinwiki-markup -> dom converter
167 2012-06-04T18:24:05 <ThomasWaldmann> it's what that converter returns
168 2012-06-04T18:24:24 <fdik> ThomasWaldmann: ah. So you're writing all parsing manually?
169 2012-06-04T18:24:36 <ThomasWaldmann> http://test.moinmo.in/+dom/Home
170 2012-06-04T18:25:10 <ThomasWaldmann> yes
171 2012-06-04T18:25:19 <fdik> ah, so your tree is some XML
172 2012-06-04T18:25:37 <ThomasWaldmann> well, it is a tree. that is the serialization of it.
173 2012-06-04T18:25:49 <fdik> ThomasWaldmann: would you be interested to have a simpler solution for that?
174 2012-06-04T18:26:01 <ThomasWaldmann> if it works :)
175 2012-06-04T18:26:19 <fdik> ThomasWaldmann: OK.
176 2012-06-04T18:26:25 <fdik> http://fdik.org/pyPEG/
177 2012-06-04T18:26:30 <fdik> That's what I'm doing.
178 2012-06-04T18:26:40 <fdik> There are other solutions, too.
179 2012-06-04T18:26:51 <fdik> (for the case you don't like what I'm doing).
180 2012-06-04T18:27:24 <fdik> With pyPEG, it would be writing a grammar, and then you have parsing, composing and convert from and to XML for free
181 2012-06-04T18:27:53 <ThomasWaldmann> what happens if the wiki page has a "syntax error"?
182 2012-06-04T18:28:06 <fdik> ThomasWaldmann: then you get a SyntaxError
183 2012-06-04T18:28:17 <fdik> ThomasWaldmann: and you can parse partially, too
184 2012-06-04T18:28:23 <ThomasWaldmann> what happens with the rest of the page?
185 2012-06-04T18:28:34 <fdik> ThomasWaldmann: that depends on how you're using it.
186 2012-06-04T18:28:50 <fdik> ThomasWaldmann: if you're parsing partially, you can resync and parse further
187 2012-06-04T18:30:08 <fdik> ThomasWaldmann: BTW: that's the project I was asking my question on #python.de for ;-)
188 2012-06-04T18:30:08 <ThomasWaldmann> sounds ok. so a wiki page with an error in the middle could look correct except the place with the error not getting transformed?
189 2012-06-04T18:30:22 <fdik> ThomasWaldmann: if you're using it that way, yes.
190 2012-06-04T18:30:54 <ThomasWaldmann> wiki users are no programmers and page transform should not fail with "syntax error on 42" :)
191 2012-06-04T18:30:56 *** dave_largo
192 2012-06-04T18:30:56 *** bilal
193 2012-06-04T18:31:06 <fdik> ThomasWaldmann: of course.
194 2012-06-04T18:31:20 <fdik> ThomasWaldmann: better: make a global exception rule, which means "all other is just text" ;-9
195 2012-06-04T18:31:32 <ThomasWaldmann> if we have a transformation x -> dom, can your tool reverse that?
196 2012-06-04T18:31:44 <fdik> ThomasWaldmann: yes, it's fully bidirectional
197 2012-06-04T18:31:59 <ThomasWaldmann> sounds cool :)
198 2012-06-04T18:32:19 <fdik> ThomasWaldmann: for a first view, please have a look on <http://fdik.org/pyPEG2/sample1.py>
199 2012-06-04T18:32:26 <ThomasWaldmann> you could make generations of soc students jobless :D
200 2012-06-04T18:32:36 <fdik> ThomasWaldmann: or just download and play around following the instructions on the homepage.
201 2012-06-04T18:32:43 <fdik> ThomasWaldmann: ;-)
202 2012-06-04T18:32:47 * fdik never intended that
203 2012-06-04T18:33:08 <ThomasWaldmann> well, not really, we have enough other stuff todo
204 2012-06-04T18:33:47 <fdik> ThomasWaldmann: It's Python 3.2 what I'm using. But for Moinmoin I will backport to Python 2.6 and support if you want.
205 2012-06-04T18:34:10 <fdik> ThomasWaldmann: pyPEG 1.x is not bidrectional, this is a feature of 2.x
206 2012-06-04T18:34:11 * ThomasWaldmann just wondered about some class def
207 2012-06-04T18:34:56 <fdik> ThomasWaldmann: the AST (what you call "DOM") is in Python objects with pyPEG
208 2012-06-04T18:35:13 <fdik> ThomasWaldmann: that's the reason for the XML export and import
209 2012-06-04T18:37:12 <ronny> oO
210 2012-06-04T18:37:23 <ronny> hmm
211 2012-06-04T18:37:29 <fdik> ThomasWaldmann: just have a look on it if you like it and tell me, please.
212 2012-06-04T18:37:31 <ThomasWaldmann> well, i just recently found that we must use polyglot generator for html5 output
213 2012-06-04T18:37:39 <fdik> ThomasWaldmann: n.p.
214 2012-06-04T18:37:58 <fdik> ThomasWaldmann: because of the XML it's a stylesheet and that's it.
215 2012-06-04T18:37:59 <ThomasWaldmann> because non-void elements are not allowed as <that /> but must be <that></that>
216 2012-06-04T18:38:15 <fdik> ThomasWaldmann: that's not a big issue
217 2012-06-04T18:38:32 <fdik> ThomasWaldmann: which component has this problem?
218 2012-06-04T18:38:49 <ronny> hmm
219 2012-06-04T18:39:03 <ThomasWaldmann> html_out, obviously :)
220 2012-06-04T18:39:38 <fdik> ThomasWaldmann: well, why not using a stylesheet for spitting out HTML? That's easy and greatly maintainable
221 2012-06-04T18:40:17 <fdik> ThomasWaldmann: maybe that's too much different to what you're doing today, I would understand that.
222 2012-06-04T18:40:56 <ronny> fdik: pyPEG looks interesting, i wonder if the ffi module could use it instead of pycparser
223 2012-06-04T18:40:58 * ThomasWaldmann is not very familiar with xml stuff. and iirc, we have no xslt processor yet.
224 2012-06-04T18:41:19 <fdik> ronny: what is the ffi module, please?
225 2012-06-04T18:41:33 <fdik> ThomasWaldmann: actually you have. Your XML library has one.
226 2012-06-04T18:41:56 <ronny> fdik: module for interfacing with c code thats way more sane than ctypes
227 2012-06-04T18:41:58 <ThomasWaldmann> emeraldtree has one?
228 2012-06-04T18:42:38 <fdik> ronny: OK.
229 2012-06-04T18:42:55 <ronny> fdik: its modeled after luas ffi
230 2012-06-04T18:44:04 <fdik> ronny: do you have a tipp for me where I can find it? Have problems with pypi about finding it
231 2012-06-04T18:44:24 <ronny> https://github.com/arigo/ffi
232 2012-06-04T18:45:46 <fdik> ronny: thx!
233 2012-06-04T18:46:25 <fdik> ronny: oh, yes, something like that can be very easy with pyPEG
234 2012-06-04T18:46:48 <ronny> hmm, but the declaratons in pyPEG look a bit painfull to read
235 2012-06-04T18:46:50 <fdik> ronny: that's one of the main targets I have, and the reasons for the defaults to parse C or C like languages
236 2012-06-04T18:46:58 <fdik> ronny: in what way, please?
237 2012-06-04T18:47:24 <ThomasWaldmann> fdik: is your stuff intended to be gpl v2 ONLY?
238 2012-06-04T18:47:39 <fdik> ThomasWaldmann: no
239 2012-06-04T18:47:59 <ThomasWaldmann> because you don't say so :)
240 2012-06-04T18:48:01 <fdik> ThomasWaldmann: I had to choose some license, and I chose one.
241 2012-06-04T18:48:05 <ronny> fdik: its strangely no-uniform
242 2012-06-04T18:48:31 <ronny> fdik: gpl v2only means gpl3 projects cant use it
243 2012-06-04T18:48:38 <fdik> ronny: ah
244 2012-06-04T18:48:44 <ronny> also ffi cant use it, since its going to be something more liberal
245 2012-06-04T18:48:45 <fdik> ronny: no, I don't want to have it like this
246 2012-06-04T18:49:15 <fdik> ronny: ? could you explain that, please?
247 2012-06-04T18:49:31 <fdik> ronny: pyPEG is somewhat unusual, because it's not only a parser
248 2012-06-04T18:49:49 <fdik> ronny: I call that "intrinsic parsing", because I didn't find something like that
249 2012-06-04T18:49:52 <ronny> fdik: explain what, why the gramar defs feel unusual?
250 2012-06-04T18:50:10 <fdik> ronny: usually, a grammar defines only the grammar of the source language, that's it
251 2012-06-04T18:50:43 <ronny> fdik: you have a weird unusual way to combin gramar and presentaton interaction at once, and even the example is non-unform
252 2012-06-04T18:50:52 <ronny> sometims its tuples, sometines its class attributes
253 2012-06-04T18:50:52 <fdik> ronny: a pyPEG grammar defines the grammar as the source language AND the substance and acidents of what's described
254 2012-06-04T18:51:28 <fdik> ronny: AND the mapping of that into the language where the AST is hold (Python in that case)
255 2012-06-04T18:51:34 <fdik> ronny: yes.
256 2012-06-04T18:51:42 <ronny> hmm
257 2012-06-04T18:51:52 <ronny> and the example c grammar is broken :)
258 2012-06-04T18:51:56 <fdik> ronny: it's a lot of theory, but you can forget that.
259 2012-06-04T18:52:09 <fdik> ronny: there is no "example C grammar", it's just a blind example ;-)
260 2012-06-04T18:52:28 <fdik> ronny: instead of see it theoretical, it's quickly to learn and then it's very easy to handle
261 2012-06-04T18:53:12 <ronny> if its non-uniform, there is more mental overhead
262 2012-06-04T18:53:29 <fdik> ronny: well, it does a lot more than a common parser does
263 2012-06-04T18:53:34 <fdik> ronny: that has to be referenced.
264 2012-06-04T18:54:02 <ronny> fdik: see zen of python?
265 2012-06-04T18:54:02 <fdik> ronny: it's more "describe the Python classes of the AST how they should be and parse in by attaching grammar rules to them".
266 2012-06-04T18:54:34 <fdik> ronny: yes
267 2012-06-04T18:55:00 <ronny> and then stuff like str subclasses
268 2012-06-04T18:55:23 <fdik> ronny: if you don't like it, n.p.
269 2012-06-04T18:55:47 * fdik will not sell it, he wrote it for himself, and published it if it's useful for someone else
270 2012-06-04T18:56:27 <ronny> hmm, it seems like it is the only parser lib availiable that can turn things back to source
271 2012-06-04T18:56:49 <fdik> ronny: yes. And this cannot be done without what you're calling "non-uniform" BTW ;-)
272 2012-06-04T18:56:55 <fdik> ronny: that's the reason
273 2012-06-04T18:57:14 <ronny> fdik: the things i call non-uniform are completely ortoonal to that feature
274 2012-06-04T18:57:15 <fdik> ronny: I'm pretty sure about that, because this is where I'm doing research in
275 2012-06-04T18:57:33 <fdik> ronny: then please help me with understanding
276 2012-06-04T18:57:47 * fdik probably does not understand what ronny means exactly
277 2012-06-04T18:58:05 <fdik> ronny: actually, pyPEG is not a PEG parser.
278 2012-06-04T18:58:21 *** MattMaker
279 2012-06-04T18:58:31 <fdik> ronny: I'm just referencing the work of Brian Ford here, because of the Packrat idea, which is from him, and he deserves it ;-)
280 2012-06-04T18:58:37 <ronny> fdik: i mean that gramar description elements are sometimes tuples, sometimes classes with attributes, sometmes strange evil builtin subclasses, sometimes subclasses of own types
281 2012-06-04T18:58:59 *** spy_
282 2012-06-04T18:59:03 <fdik> ronny: yes, and unfortunately for each of them there is a good reason.
283 2012-06-04T18:59:21 <fdik> ronny: of course, for the "evil subclasses of str" I could write wrappers
284 2012-06-04T18:59:30 * fdik hates wrappers
285 2012-06-04T19:00:27 <fdik> ronny: perhaps I will do that for Symbol
286 2012-06-04T19:01:05 <ronny> i'd appreciate a more pythonic initial feel
287 2012-06-04T19:01:11 <fdik> ronny: I just didn't, because Symbol has to be an exception anyways
288 2012-06-04T19:01:21 <fdik> ronny: whatever you mean here.
289 2012-06-04T19:01:22 <ronny> its on my dig into after thesis list now
290 2012-06-04T19:01:32 <fdik> ronny: ;-)
291 2012-06-04T19:01:51 <fdik> ronny: if you have questions, I'd be pleased to explain.
292 2012-06-04T19:02:02 <ronny> fdik: basically i dont like that there are so many ways to spell gramar elements, and once i dig i'll probably find more things
293 2012-06-04T19:02:18 <fdik> ronny: I'd like to have feedback - I'm pretty sure I can explain the most, and I'm as sure that you can show me my mistakes ;-)
294 2012-06-04T19:02:38 <fdik> ronny: it's NOT "grammar elements". That's the first problem.
295 2012-06-04T19:02:45 <fdik> ronny: this is not a common grammar.
296 2012-06-04T19:03:08 <fdik> ronny: it's the structure of the AST as real classes and objects, and their interrelation with grammar
297 2012-06-04T19:03:21 <fdik> ronny: it cannot be done without
298 2012-06-04T19:03:30 <ronny> fdik: for me thats not different from a gramar
299 2012-06-04T19:03:31 <fdik> ronny: that's the reason why it can compose
300 2012-06-04T19:03:39 <fdik> ronny: you're mistaken here, I think
301 2012-06-04T19:04:00 <fdik> ronny: a common AST is a typed tree
302 2012-06-04T19:05:05 <fdik> ronny: what pyPEG's parse() spits out is something else - it's objects with attributes and own different structure, where you can find the AST as part of this structure
303 2012-06-04T19:05:22 <fdik> ronny: it's even cyclic BTW, so I'm using weakref
304 2012-06-04T19:05:39 <ronny> so it spits out something thats a ast + the st for reversal?
305 2012-06-04T19:05:58 <fdik> ronny: it spits out what's directly described
306 2012-06-04T19:06:18 <fdik> ronny: and it knows about languages inside languages and their connection using names
307 2012-06-04T19:06:56 <fdik> ronny: following a transformation of the theory in Kripke's "Naming and Necessity" to a positivistic model
308 2012-06-04T19:07:23 <ronny> i dont know that one
309 2012-06-04T19:07:34 <fdik> ronny: and it's "if you're using it this way, it will do what you expect" ;-)
310 2012-06-04T19:08:18 <ronny> km, definitivel something for afte the thesis, its getting to the too much 3rd party input requird state already
311 2012-06-04T19:08:20 <fdik> ronny: a huge part of my work is "how can I find good defaults and move it in way that people who don't understand a word of the theory and are not interested in can use it naturally"
312 2012-06-04T19:08:43 <fdik> ronny: I didn't publish BTW, sorry.
313 2012-06-04T19:09:10 <fdik> ronny: pyPEG 1 was just a prototype, and when I saw that it's useful itsel, I published the software (not the theory behind)
314 2012-06-04T19:09:34 <fdik> ronny: pyPEG 2 is what I'm using now myself, because I kicked away C++ and writing things in Python now, not only the prototype ;-)
315 2012-06-04T19:10:30 <fdik> ronny: what are you working in?
316 2012-06-04T19:10:37 * fdik is just curious
317 2012-06-04T19:11:04 <ronny> im buiding a distributed ci on top of a distributed database, trying a few new concepts
318 2012-06-04T19:11:17 <fdik> ronny: what is a "ci", please?
319 2012-06-04T19:11:34 <ronny> continious integration
320 2012-06-04T19:12:16 <fdik> ronny: sounds interesting
321 2012-06-04T19:13:09 <ronny> fdik: it is, distributed statemachines are fun
322 2012-06-04T19:13:12 * fdik is interested in distributed systems and automating software development, and that sounds like the two of them ;-)
323 2012-06-04T19:13:52 <ronny> oh, btw - we accept complaints about hijacking the channel :
324 2012-06-04T19:13:53 <fdik> ronny: yes, and solving the coherency problem without being to synchronous is a very interesting topic ;-)
325 2012-06-04T19:14:07 <fdik> ronny: *oops* you're right
326 2012-06-04T19:14:23 * fdik becomes red
327 2012-06-04T19:15:41 <ronny> fdik: the basic gist is, that building things on top of a database with master master replicaton and mvcc solves a lot of the lowlevel issues :P
328 2012-06-04T19:17:21 <fdik> ronny: I beleive that.
329 2012-06-04T19:17:40 <fdik> ronny: nice topic
330 2012-06-04T19:17:45 <fdik> ronny: really :-))
331 2012-06-04T19:17:54 *** jaiditya
332 2012-06-04T19:18:12 *** jaiditya
333 2012-06-04T19:18:36 <fdik> ronny: I would be pleased to read your thesis when it's ready (if you can't talk about details before, I understand)
334 2012-06-04T19:20:03 <ronny> i can talk, but i think i need to do some back to working :P
335 2012-06-04T19:20:22 <ronny> but feedback of interested parties is helfull :P
336 2012-06-04T19:23:41 *** greg_f
337 2012-06-04T19:25:17 <fdik> ronny: I'm *very* interested.
338 2012-06-04T19:25:38 <fdik> ronny: BTW: do you a way to get a value semantics with Python without subclassing a basic type?
339 2012-06-04T19:25:54 <fdik> ronny: that's the reason for class Symbol(str)
340 2012-06-04T19:26:57 <fdik> ronny: and I don't like the solution, too, because of the clashing with metaprogramming because of the different layout
341 2012-06-04T19:27:15 <ronny> fdik: 'value semantics'
342 2012-06-04T19:27:23 <fdik> ronny: ah, you're on #python.de, I'm asking there ;-)
343 2012-06-04T20:12:57 *** Oglarun
344 2012-06-04T20:14:00 *** spy_
345 2012-06-04T20:15:21 *** spy_
346 2012-06-04T20:47:17 *** spy_
347 2012-06-04T20:49:12 *** spy_
348 2012-06-04T21:01:00 *** spy
349 2012-06-04T21:05:51 *** spy_
350 2012-06-04T22:12:08 *** fdik
351 2012-06-04T22:32:58 <dreimark> moin
352 2012-06-04T23:03:15 *** dwcramer
353 2012-06-04T23:15:09 *** dave_largo
354 2012-06-04T23:54:33 *** MattMaker
355
