| 00:47.04 | *** join/#storm darkstar_ (n=nobody@198.Red-213-98-22.staticIP.rima-tde.net) |
| 00:47.14 | darkstar_ | hello everybody |
| 00:47.28 | darkstar_ | please, I'm becoming crazy because of storm and datetime |
| 00:47.55 | darkstar_ | I don't know why I can't set current time (hour, minute and second included) |
| 00:48.12 | darkstar_ | if anybody could help my I'll be his slave for ever :P |
| 00:48.52 | darkstar_ | I'm using datetime.datetime.now() to set the current date |
| 00:49.04 | *** part/#storm exarkun (n=exarkun@wordeology.com) |
| 00:50.19 | darkstar_ | I allways get in the database 2009-09-19 00:00:00 |
| 00:50.58 | wgrant | darkstar_: That works fine for me. Sure your types are correct in the DB schema and Storm class definition? |
| 00:51.10 | darkstar_ | I think so |
| 00:51.19 | darkstar_ | I've datetime in database (mysql) |
| 00:51.37 | darkstar_ | and Date() in storm |
| 00:51.48 | darkstar_ | is this correct? |
| 00:54.41 | darkstar_ | omg! I think it should be DateTime() |
| 00:55.01 | darkstar_ | yessssssss |
| 00:55.24 | darkstar_ | I've spent half an hour with it! |
| 00:55.25 | darkstar_ | hehe |
| 00:55.30 | darkstar_ | thanks you very much, wgrant |
| 00:55.35 | darkstar_ | nice to meet you |
| 00:55.48 | wgrant | darkstar_: Haha, glad it was easy. |
| 00:55.59 | darkstar_ | hehe |
| 00:56.12 | darkstar_ | this is only one problem |
| 00:56.28 | darkstar_ | now I've another more difficult with set |
| 00:57.10 | darkstar_ | I've to set a few values and some of them are references |
| 00:57.30 | darkstar_ | so I think it should be like... |
| 00:59.33 | darkstar_ | mmm, now it seems works... |
| 01:00.22 | darkstar_ | here is the problem |
| 01:00.31 | darkstar_ | AttributeError: 'Reference' object has no attribute 'variable_factory' |
| 01:02.36 | darkstar_ | I don't know what's the problem, because I can do it what I'm trying with a new object |
| 01:02.57 | darkstar_ | but when I do it in set... it fails |
| 01:04.42 | darkstar_ | getArticle(articulo["ref"], True).set(titulo = unicode(articulo["titulo"]), contenido = unicode(articulo["contenido"]), ultima_edicion = datetime.datetime.now(), estado = unicode(articulo["estado"]), ocupado = 0, tema = store.get(frikisdb.Tema, find("Tema", {"nombre": articulo["tema"]}, True).one().id_tema).id_tema) |
| 01:04.53 | darkstar_ | problem is here |
| 01:05.02 | darkstar_ | tema = store.get(frikisdb.Tema, find("Tema", {"nombre": articulo["tema"]}, True).one().id_tema).id_tema |
| 01:12.09 | darkstar_ | wgrant, how can I update a field that is a reference using set? |
| 01:13.02 | wgrant | darkstar_: I don't know. Why not just set it directly? |
| 01:13.22 | darkstar_ | 'cause then it doesn't update |
| 01:13.42 | darkstar_ | if it's on the database |
| 01:13.43 | wgrant | It certainly should. |
| 01:13.48 | darkstar_ | I can't add it to store |
| 01:14.07 | darkstar_ | i've been trying it, wgrant |
| 01:14.14 | darkstar_ | if yo could tell me how... |
| 01:14.32 | darkstar_ | store.add(object) isn't allowed if it's on database |
| 01:14.38 | darkstar_ | store.flush() do nothing with it |
| 01:14.42 | darkstar_ | store.commit() same |
| 01:14.51 | darkstar_ | so how do you do it? |
| 01:15.26 | wgrant | darkstar_: What do you mean by "isn't allowed if it's on the database"? You mean you can't add it if it's already there? |
| 01:15.39 | darkstar_ | I meant this |
| 01:15.46 | wgrant | You don't need to. |
| 01:15.50 | wgrant | Just set the attributes. |
| 01:15.55 | darkstar_ | I've few notices |
| 01:16.00 | darkstar_ | if I want to edit one |
| 01:16.08 | darkstar_ | I find it in the table |
| 01:16.10 | wgrant | When you flush or commit, the changes will end up in the DB. |
| 01:16.16 | darkstar_ | get it on an object |
| 01:16.25 | darkstar_ | and now I want to change the values of the fields |
| 01:16.42 | darkstar_ | I wish, wgrant |
| 01:16.46 | darkstar_ | but it doesn't work |
| 01:16.54 | darkstar_ | I'm trying right now |
| 01:17.10 | wgrant | darkstar_: What does Store.of(thatobject) say? |
| 01:17.18 | darkstar_ | article = getArticle(articulo["ref"], True).one() |
| 01:17.25 | darkstar_ | article.tema = store.get(frikisdb.Tema, find("Tema", {"nombre": articulo["tema"]}, True).one().id_tema).id_tema |
| 01:17.35 | darkstar_ | store.commit() |
| 01:17.41 | darkstar_ | and nothing change in my db |
| 01:17.53 | darkstar_ | I'll tell you right know |
| 01:17.58 | wgrant | darkstar_: Article.tema is a Reference? |
| 01:18.03 | darkstar_ | yes |
| 01:18.11 | darkstar_ | it is |
| 01:18.16 | wgrant | darkstar_: What is id_tema? It sounds like it is the integer primary key. |
| 01:18.20 | darkstar_ | a reference to another table called Tema |
| 01:18.22 | darkstar_ | yes |
| 01:18.27 | darkstar_ | id_tema es the primary key |
| 01:18.27 | wgrant | OK, that could be your problem. |
| 01:18.36 | wgrant | You assign the object to a reference, not the object's primary key. |
| 01:18.44 | darkstar_ | so... |
| 01:18.48 | wgrant | So maybe just article.tema = frikisdb.Tema, find("Tema", {"nombre": articulo["tema"]}, True).one() |
| 01:18.55 | wgrant | Er. |
| 01:19.08 | darkstar_ | I think it fail |
| 01:19.08 | wgrant | Not quite that. I didn't read deeply enough. |
| 01:19.17 | darkstar_ | but I'll try it now |
| 01:19.20 | wgrant | What is this 'find' function? |
| 01:19.38 | wgrant | Why are you getting the ID, looking it up in the store, then getting the ID out again? |
| 01:20.14 | darkstar_ | storm.exceptions.WrongStoreError: <frikisdb.Articulo object at 0xb76ef90c> and <frikisdb.Tema object at 0xb76f224c> cannot be linked because they are in different stores. |
| 01:20.25 | darkstar_ | <PROTECTED> |
| 01:20.25 | darkstar_ | <PROTECTED> |
| 01:20.30 | wgrant | darkstar_: OK, this is looking better. But why do you have multiple stores? |
| 01:20.50 | darkstar_ | I'm trying to update an article |
| 01:20.55 | darkstar_ | article has few references |
| 01:21.17 | wgrant | Right. |
| 01:21.24 | darkstar_ | I don't know why there is no problem if I set it with equal |
| 01:21.25 | wgrant | But there seem to be at least two stores in use. |
| 01:21.33 | wgrant | There should probably only be one. |
| 01:21.34 | darkstar_ | but it doesn't work with set |
| 01:21.40 | darkstar_ | no |
| 01:21.42 | darkstar_ | just one |
| 01:21.54 | darkstar_ | <PROTECTED> |
| 01:21.54 | darkstar_ | <PROTECTED> |
| 01:21.54 | darkstar_ | <PROTECTED> |
| 01:21.54 | darkstar_ | <PROTECTED> |
| 01:21.54 | darkstar_ | <PROTECTED> |
| 01:21.56 | wgrant | That error suggests says there are at least two. |
| 01:22.17 | darkstar_ | error can suggest whatever he want |
| 01:22.18 | wgrant | That last line doesn't make sense to me. What is it doing? |
| 01:22.27 | darkstar_ | what? |
| 01:22.32 | darkstar_ | uf |
| 01:22.40 | darkstar_ | what I told you few lines before |
| 01:22.45 | darkstar_ | changing the reference |
| 01:22.53 | darkstar_ | because I want to change this field |
| 01:22.58 | darkstar_ | and this field is a reference |
| 01:23.15 | darkstar_ | I just want to change it |
| 01:23.20 | darkstar_ | and it seems impossible |
| 01:23.35 | wgrant | What is this 'find' thing? |
| 01:23.53 | wgrant | And why are you finding an object, taking its ID, looking it up in (probably the wrong) store, then taking the ID? |
| 01:23.54 | darkstar_ | find sentence works |
| 01:24.13 | wgrant | What does it return? |
| 01:24.13 | darkstar_ | I haven't the id |
| 01:24.17 | darkstar_ | I need the id |
| 01:24.21 | wgrant | Why? |
| 01:24.27 | wgrant | The reference doesn't need the ID. |
| 01:24.32 | darkstar_ | first of all I only have the mame |
| 01:24.35 | wgrant | The reference needs the actual object. |
| 01:24.44 | darkstar_ | again |
| 01:24.54 | darkstar_ | have you read what I posted before? |
| 01:24.55 | wgrant | article.tema = store.find(frikisdb.Tema, name='something').one()? |
| 01:25.02 | darkstar_ | <darkstar_> storm.exceptions.WrongStoreError: <frikisdb.Articulo object at 0xb76ef90c> and <frikisdb.Tema object at 0xb76f224c> cannot be linked because they are in different stores. |
| 01:25.02 | darkstar_ | <darkstar_> File "dbmanager.py", line 85, in addArticle |
| 01:25.02 | darkstar_ | <darkstar_> article.tema = store.get(frikisdb.Tema, find("Tema", {"nombre": articulo["tema"]}, True).one().id_tema) |
| 01:25.15 | darkstar_ | if I do what you said, it fails |
| 01:25.15 | wgrant | OK, try: |
| 01:25.27 | wgrant | article.team = Store.of(article).find(frikisdb.Tema, name='something').one() |
| 01:26.17 | wgrant | That error means what it says. You've got a second store floating around somewhere. That last statement I gave should use the same store. |
| 01:27.31 | darkstar_ | and nothing happends |
| 01:27.46 | darkstar_ | should I do a commit before that? |
| 01:27.50 | darkstar_ | commit, flush... |
| 01:28.00 | darkstar_ | nothing seems to work |
| 01:28.18 | wgrant | The changes are in a transaction, so you need to commit before you can see them in any other transaction. |
| 01:28.50 | darkstar_ | ok |
| 01:28.53 | darkstar_ | I've commited |
| 01:28.57 | darkstar_ | nothing changes |
| 01:29.16 | wgrant | Did you fix the typo I made? |
| 01:29.33 | darkstar_ | article.tema = Store.of(article).find(frikisdb.Tema, nombre = unicode(articulo["tema"])).one() |
| 01:30.09 | wgrant | And the value in the database does not change? What if you examine article.tema afterwards from within the program? |
| 01:30.39 | darkstar_ | I'll tell you now |
| 01:31.07 | wgrant | Also, check from within the program the integer value that article.tema references. |
| 01:31.23 | darkstar_ | well, if I print it I get a reference to an object |
| 01:31.30 | darkstar_ | <frikisdb.Tema object at 0xb76b8ecc> |
| 01:31.37 | darkstar_ | but wich... I don't know |
| 01:31.52 | wgrant | darkstar_: What is the line that defines the Article.tema Reference? |
| 01:32.18 | darkstar_ | <PROTECTED> |
| 01:32.22 | darkstar_ | <PROTECTED> |
| 01:32.27 | darkstar_ | in class Articulo(Storm): |
| 01:32.56 | wgrant | OK. What is article.id_tema after you set article.tema, and how are you committing? |
| 01:34.09 | darkstar_ | value is correct |
| 01:34.10 | darkstar_ | 2 |
| 01:34.13 | wgrant | Great. |
| 01:34.17 | darkstar_ | the value I whant to change to |
| 01:34.19 | wgrant | So, you're probably not committing properly. |
| 01:34.20 | darkstar_ | I'm commiting |
| 01:34.26 | darkstar_ | <PROTECTED> |
| 01:34.26 | darkstar_ | <PROTECTED> |
| 01:34.26 | darkstar_ | <PROTECTED> |
| 01:34.27 | wgrant | This comes back to the multiple store issue. |
| 01:34.28 | darkstar_ | and |
| 01:34.33 | wgrant | Try Store.of(article).commit() |
| 01:34.45 | darkstar_ | ok |
| 01:35.15 | darkstar_ | eureka |
| 01:35.30 | wgrant | Excellent. |
| 01:35.38 | darkstar_ | gloooooria gloria aleluuuuuuuya |
| 01:35.39 | darkstar_ | jajajaja |
| 01:35.40 | wgrant | So, most of your problems are because you have a second store, somehow. |
| 01:35.44 | darkstar_ | thanks you very much, wgrant !!!! |
| 01:35.48 | wgrant | np |
| 01:35.51 | darkstar_ | mmmmm |
| 01:35.53 | wgrant | Now track down and get rid of that store! |
| 01:36.23 | darkstar_ | but I don't know where can I have this store |
| 01:36.35 | darkstar_ | mmmm |
| 01:36.37 | wgrant | darkstar_: Where are you getting 'article' from? |
| 01:36.50 | darkstar_ | maybe I should close... |
| 01:36.55 | darkstar_ | store.close() ? |
| 01:37.05 | wgrant | No. |
| 01:37.23 | wgrant | You are getting 'article' from somewhere other than 'store'. |
| 01:37.32 | wgrant | Or you're redefining 'store' somewhere. |
| 01:37.38 | darkstar_ | mabe on other functions |
| 01:37.53 | darkstar_ | functions called throw this one |
| 01:38.03 | darkstar_ | should I do anything? |
| 01:38.49 | wgrant | You'll have to hunt through your code and work out what's going bad. I can't help you with that side of things |
| 01:39.06 | darkstar_ | yes, I know |
| 01:39.26 | darkstar_ | you have help me more than I belive anybody will help me |
| 01:39.37 | darkstar_ | so thanks you very much because I was becomming crazy |
| 01:39.43 | wgrant | Heh. |
| 01:39.44 | wgrant | No problem. |
| 01:39.57 | *** join/#storm sidnei (n=sidnei@201.1.232.244) |
| 01:40.13 | darkstar_ | wgrant, storm and mod_python will kill me |
| 01:40.25 | wgrant | darkstar_: mod_python will kill you. Storm will not! |
| 01:40.31 | darkstar_ | mod_python is horrible |
| 01:40.44 | wgrant | Why not use WSGI? |
| 01:40.53 | sidnei | yeah, mod_wsgi ftw! |
| 01:40.56 | darkstar_ | because I don't know what is it |
| 01:41.42 | wgrant | WSGI is awesome. |
| 01:41.50 | wgrant | And will hopefully eliminate mod_python soon. |
| 01:42.12 | wgrant | It defines a standard Python web server interface. |
| 01:42.15 | wgrant | So you're not tied to Apache, for example. |
| 01:42.30 | wgrant | And the interface is a bit less awful than mod_python. |
| 01:42.59 | darkstar_ | mod_python major problem is when something fails |
| 01:43.04 | darkstar_ | you can get old results |
| 01:43.26 | darkstar_ | but now I've my own htmlhandler and everything done |
| 01:43.45 | darkstar_ | if I use wsgi I'll lose all the done job |
| 01:45.12 | darkstar_ | omg, it seems quite easier and much better :'( |
| 01:45.30 | wgrant | That's what everybody says. |
| 01:45.37 | wgrant | I don't know why more don't know about it :( |
| 01:45.44 | darkstar_ | I'll cry |
| 01:46.05 | darkstar_ | do you have a time machine? |
| 01:46.41 | wgrant | brb, I'll fetch it. |
| 01:50.45 | darkstar_ | now mod_python is f******* me |
| 01:50.49 | darkstar_ | :'( |
| 01:51.14 | darkstar_ | I need to think in other thing or I'll break my keyboard |
| 02:15.15 | jkakar | wgrant is right, WSGI is the way forward. |
| 02:15.32 | jkakar | I've used modpython and it only leads to pain. |
| 02:17.02 | wgrant | +inf |
| 02:56.25 | darkstar_ | wgrant, do you know a different type for strings? |
| 02:56.28 | darkstar_ | instead unicode |
| 02:56.38 | *** join/#storm keppla (n=keppla@i577B35DB.versanet.de) |
| 03:20.31 | darkstar_ | wgrant, do you know a different type for string that allow iso-8859-15? I need tildes & co. |
| 03:28.56 | lifeless | darkstar_: use unicode. it can do tildes etc. |
| 03:29.09 | darkstar_ | I can't |
| 03:29.13 | darkstar_ | looki |
| 03:30.42 | darkstar_ | TypeError: Expected unicode, found <type 'str'>: 'P\xe1nico' |
| 03:31.06 | darkstar_ | and if I try to unicode this word |
| 03:31.16 | darkstar_ | unicode("Pánico") |
| 03:31.20 | darkstar_ | UnicodeDecodeError: 'ascii' codec can't decode byte 0xe1 in position 1: ordinal not in range(128) |
| 03:33.29 | lifeless | >>> u"ú" |
| 03:33.29 | lifeless | u'\xfa' |
| 03:33.41 | lifeless | anyhow, this is a #python issue really |
| 03:33.46 | lifeless | there are docs and faqs. |
| 03:33.51 | darkstar_ | yes |
| 03:33.55 | darkstar_ | but what about variables? |
| 03:34.06 | darkstar_ | I can't do u"var" |
| 03:34.13 | lifeless | Strings you type that python parses are in an encoding already. |
| 03:34.17 | lifeless | yes you can ;) |
| 03:34.23 | lifeless | <PROTECTED> |
| 03:34.32 | darkstar_ | >>> a |
| 03:34.32 | darkstar_ | 'P\xe1nico' |
| 03:34.32 | darkstar_ | >>> u"a" |
| 03:34.32 | darkstar_ | u'a' |
| 03:34.32 | lifeless | if you mean variable names, then no. |
| 03:34.35 | darkstar_ | that is the problem |
| 03:35.43 | lifeless | do you know about __repr__ ? |
| 03:35.48 | lifeless | when you do >>> FOO |
| 03:35.54 | lifeless | FOO is _not_ shown |
| 03:35.58 | lifeless | repr(FOO) is shown |
| 03:36.19 | lifeless | print u"\xe1" |
| 03:36.39 | lifeless | I encourage you to read the python docs on str and unicode, console encodings and implicit conversions. |
| 03:36.46 | lifeless | then you will have some idea about what is going on |
| 03:36.51 | darkstar_ | I don't know what are you talking about |
| 03:36.58 | darkstar_ | do you know how to do it or not? |
| 03:37.05 | darkstar_ | I've read few books |
| 03:37.24 | darkstar_ | and I know how to change from utf to iso |
| 03:37.30 | darkstar_ | but from iso to utf I don't know |
| 03:37.36 | darkstar_ | I'm looking for it and I get nothing |
| 03:37.51 | lifeless | variable.decode('iso-8859-15?') |
| 03:37.52 | darkstar_ | if there are a tilde or something like that... |
| 03:38.06 | darkstar_ | exactly, with decode and encode |
| 03:38.13 | darkstar_ | .decode('utf8').encode('iso-8859-15') |
| 03:38.21 | darkstar_ | is from utf8 to iso |
| 03:38.32 | darkstar_ | but if I want to do the opposite |
| 03:39.04 | darkstar_ | what the hell |
| 03:39.09 | darkstar_ | wait |
| 03:39.42 | darkstar_ | UnicodeEncodeError: 'ascii' codec can't encode character u'\xf1' in position 2: ordinal not in range(128) |
| 03:39.53 | lifeless | thats an implicit conversion happenin |
| 03:39.54 | darkstar_ | <PROTECTED> |
| 03:39.54 | darkstar_ | <PROTECTED> |
| 03:39.55 | lifeless | g |
| 03:40.17 | lifeless | tema.nombre will be unicode |
| 03:40.51 | darkstar_ | tema.nombre has a tilde |
| 03:40.54 | lifeless | you shouldn't decode a unicode string; decode only makes sense on the 'str' object type, and encode only makes sense on the 'unicode' object type |
| 03:41.05 | darkstar_ | tema.nombre has a tilde |
| 03:41.08 | lifeless | so? |
| 03:41.11 | darkstar_ | if I don't do this |
| 03:41.30 | darkstar_ | <PROTECTED> |
| 03:41.32 | darkstar_ | UnicodeEncodeError: 'ascii' codec can't encode character u'\xf1' in position 2: ordinal not in range(128) |
| 03:42.01 | lifeless | find where that is coming from |
| 03:42.13 | darkstar_ | I know it |
| 03:42.15 | lifeless | is your DB provider not returning unicode, for instance |
| 03:42.16 | darkstar_ | there are 3 "temas" |
| 03:42.23 | darkstar_ | of course not |
| 03:42.28 | darkstar_ | it's using latin1 |
| 03:42.34 | darkstar_ | ant it's what I want |
| 03:42.41 | darkstar_ | could I still using latin1? |
| 03:42.49 | lifeless | I think you misunderstand my question |
| 03:43.02 | lifeless | I did not say 'what encoding the database is using' |
| 03:43.12 | lifeless | I said 'is the DB provider not returning unicode' |
| 03:43.18 | lifeless | unicode is an object type, not an encoding. |
| 03:44.01 | darkstar_ | so... |
| 03:44.25 | lifeless | so something is causing an implicit ascii conversion in your list comprehension |
| 03:44.57 | lifeless | implicit conversions will fail in many different ways |
| 03:45.19 | lifeless | python 3K is much better at this |
| 03:45.32 | darkstar_ | so what can I do? |
| 03:46.07 | lifeless | make sure your DB provider is returning unicode, as I said above |
| 03:46.21 | lifeless | also, if you aren't tracking whats going on, read up as I said even earlier. |
| 03:46.40 | darkstar_ | read about what |
| 03:46.46 | lifeless | this is a _very specific_ python issue |
| 03:46.53 | lifeless | 13:35 < lifeless> I encourage you to read the python docs on str and unicode, console encodings and implicit conversions. |
| 03:46.54 | darkstar_ | I've told you that I'm using latin1 on my db |
| 03:47.03 | lifeless | sigh, whatever. |
| 03:47.06 | darkstar_ | and I've ask you if this could be the problem |
| 03:47.12 | darkstar_ | and if it is the problem |
| 03:47.24 | darkstar_ | how could I solve it and still using latin1 |
| 03:47.29 | darkstar_ | and if not |
| 03:47.30 | wgrant | darkstar_: What are you doing with the result of that function? |
| 03:47.37 | darkstar_ | I'll change it to another codification |
| 03:47.50 | lifeless | its not the use of latin1 in the DB |
| 03:47.54 | lifeless | its configuration or code bugs |
| 03:48.14 | lifeless | I can't tell you where they are, because I don't have your code. The backtrace should be a good place to start debugging |
| 03:48.33 | darkstar_ | show it in html |
| 03:48.42 | lifeless | but again, if you don't actually understand what python is doing with encodings and str and unicode objects, you'll find it quite hard. Which is why I'm trying to get you to educate yourself |
| 03:48.42 | darkstar_ | inside a formulary |
| 03:49.23 | wgrant | darkstar_: How are you getting it into the HTML? |
| 03:49.55 | darkstar_ | formulario.option(tuple(dbmanager.getTopics())) |
| 03:50.06 | darkstar_ | function getTopics is the one who fails |
| 03:50.14 | darkstar_ | formulario.option |
| 03:50.29 | darkstar_ | just do this |
| 03:50.48 | darkstar_ | for i in options: |
| 03:50.49 | darkstar_ | <PROTECTED> |
| 03:50.49 | darkstar_ | <PROTECTED> |
| 03:50.49 | wgrant | formulario.option is very probably not unicode-aware, and is trying to coerce its input to a str. |
| 03:50.49 | darkstar_ | ... |
| 03:50.51 | darkstar_ | no more |
| 03:51.17 | wgrant | Hmm. |
| 03:51.48 | darkstar_ | it's just for bild a option/value form |
| 03:52.00 | wgrant | Which line is throwing the UnicodeEncodeError? |
| 03:53.22 | darkstar_ | <PROTECTED> |
| 03:53.22 | darkstar_ | UnicodeEncodeError: 'ascii' codec can't encode character u'\xf1' in position 584: ordinal not in range(128) |
| 03:53.30 | wgrant | Aha. |
| 03:53.43 | darkstar_ | now is another... I hate mod_python |
| 03:53.43 | wgrant | That's not the error you showed before. |
| 03:53.45 | wgrant | This makes more sense. |
| 03:53.51 | wgrant | This isn't mod_python's fault. |
| 03:53.56 | darkstar_ | before was another one |
| 03:54.01 | darkstar_ | but mod_python... |
| 03:54.18 | lifeless | so req is an ascii |
| 03:54.20 | lifeless | stream |
| 03:54.22 | lifeless | content is unicode |
| 03:54.22 | wgrant | You're trying to write unencoded Unicode data to a bytestream. |
| 03:54.28 | lifeless | you have to encode |
| 03:54.43 | wgrant | So you need to pick an encoding (hopefully utf-8), specify it in your HTTP headers, and encode the unicode string. |
| 03:55.12 | darkstar_ | but I'd like to use latin1 for clients |
| 03:55.22 | wgrant | Use latin1, then. |
| 03:55.35 | darkstar_ | thats what I'm using now |
| 03:55.39 | lifeless | [but using anything other than utf8 is totally annoying for web pages] |
| 03:55.55 | wgrant | Right, I strongly suggest that everybody use UTF-8 |
| 03:56.18 | darkstar_ | the problem with utf8 is that people who use iso can't see it very well |
| 03:56.20 | wgrant | darkstar_: So just encode the content as latin-1 before you write it outl |
| 03:56.33 | darkstar_ | mmm. I'll try |
| 03:56.34 | wgrant | Um, are there really non-UTF-8 clients out there? |
| 03:56.43 | wgrant | s/clients/browsers/ |
| 03:56.53 | wgrant | ...... wtf? |
| 03:57.03 | lifeless | darkstar_: what do you mean 'people who use iso can't see utf8 very well' ? |
| 03:57.06 | wgrant | I like it. |
| 03:58.35 | darkstar_ | oh, shit |
| 03:58.39 | darkstar_ | now the problem is for send it |
| 03:58.42 | darkstar_ | :-( |
| 03:59.07 | darkstar_ | lifeless, yes |
| 03:59.12 | darkstar_ | I use iso |
| 03:59.17 | darkstar_ | iso is the standard |
| 03:59.34 | wgrant | Not really. |
| 03:59.45 | darkstar_ | well, it's an ISO |
| 03:59.46 | darkstar_ | hehe |
| 03:59.48 | wgrant | I think just about everybody except Microsoft has standardised on UTF-8 as the universal encoding. |
| 04:00.21 | darkstar_ | I'm a nostalgic |
| 04:00.26 | darkstar_ | hehehe |
| 04:00.37 | darkstar_ | I'll use utf8 when everybody use it |
| 04:00.50 | darkstar_ | but there are people who still using iso |
| 04:01.10 | lifeless | darkstar_: its a lot easier for /everyone/ if you use UTF-8 |
| 04:02.01 | darkstar_ | I should change everything |
| 04:02.11 | darkstar_ | whats the name for html headers? |
| 04:02.15 | darkstar_ | utf8 ? |
| 04:02.19 | lifeless | MSIE, firefox, opera and chrome all support utf8 |
| 04:02.23 | darkstar_ | or UTF-8 ? |
| 04:02.36 | wgrant | utf-8 |
| 04:02.37 | lifeless | do you mean the content encoding value? |
| 04:02.43 | darkstar_ | yes |
| 04:02.50 | darkstar_ | and python encoding for header too |
| 04:02.58 | darkstar_ | I've never use utf8 before |
| 04:03.00 | lifeless | what python encoding? |
| 04:03.28 | darkstar_ | # -*- coding: iso-8859-15 -*- |
| 04:03.35 | darkstar_ | I've this in all my python files |
| 04:03.37 | lifeless | thats irrelevant to your web page |
| 04:03.48 | lifeless | you can keep using that if you want, noone will care |
| 04:04.26 | darkstar_ | content="text/html;charset=utf-8" |
| 04:04.28 | darkstar_ | like this? |
| 04:04.47 | wgrant | Yes. |
| 04:05.55 | darkstar_ | ok, I'll try |
| 04:05.56 | lifeless | http://www.iana.org/assignments/character-sets |
| 04:06.00 | lifeless | Name: UTF-8 [RFC3629] |
| 04:06.43 | lifeless | btw, UTF-8 encodes unicode, which is ISO 10646 |
| 04:06.48 | lifeless | so its also a standard ;) |
| 04:06.58 | darkstar_ | brrrr |
| 04:07.00 | darkstar_ | heheehe |
| 04:22.00 | darkstar_ | lifeless, same error |
| 04:24.26 | darkstar_ | <PROTECTED> |
| 04:24.27 | darkstar_ | UnicodeDecodeError: 'ascii' codec can't decode byte 0xf3 in position 1210: ordinal not in range(128) |
| 04:24.37 | darkstar_ | and now everything is utf8 |
| 04:59.48 | darkstar_ | <PROTECTED> |
| 04:59.48 | darkstar_ | <PROTECTED> |
| 04:59.48 | darkstar_ | UnicodeEncodeError: 'ascii' codec can't encode character u'\xf1' in position 598: ordinal not in range(128) |
| 04:59.53 | darkstar_ | :'( |
| 05:00.01 | darkstar_ | utf8 hasn't help me |
| 05:00.52 | wgrant | darkstar_: That means you haven't encoded it yet... |
| 05:01.19 | darkstar_ | everything is encoded tu utf8 |
| 05:01.35 | wgrant | Apparently no, as you're trying to write out a unicode object. |
| 05:01.38 | wgrant | s/no/not/ |
| 05:02.59 | darkstar_ | now don't work things works before |
| 05:03.05 | darkstar_ | doesn't work |
| 05:05.04 | wgrant | You probably accidentally turned content back into a unicode after encoding it. |
| 05:05.39 | darkstar_ | turned back into unicode? |
| 05:05.45 | darkstar_ | but is unicode what I need |
| 05:06.10 | wgrant | Turned it into *a* unicode, not turned it into Unicode. |
| 05:06.21 | wgrant | "a unicode" meaning a Python object of type 'unicode' |
| 05:06.32 | darkstar_ | I'm becoming crazy, wgrant |
| 05:06.37 | darkstar_ | everything is in unicode |
| 05:06.41 | darkstar_ | and the error is |
| 05:06.50 | darkstar_ | UnicodeEncodeError: 'ascii' codec can't encode character u'\xf1' in position 584: ordinal not in range(128) |
| 05:07.03 | wgrant | Right. |
| 05:07.06 | darkstar_ | wth |
| 05:07.30 | wgrant | You need to give req.write a str (that is to say a bytestring) encoded as utf-8. |
| 05:07.49 | darkstar_ | whell |
| 05:07.49 | wgrant | You get that by taking a Python unicode object (which has no encoding), and calling .encod('utf-8') on it. |
| 05:07.56 | darkstar_ | if I do |
| 05:08.03 | darkstar_ | req.write(unicode(content)) |
| 05:08.09 | darkstar_ | I get same error |
| 05:08.22 | wgrant | Of course. |
| 05:08.31 | wgrant | req.write(content.encode('utf-8')) |
| 05:08.49 | wgrant | How I wish Python 3 was universal already... |
| 05:09.04 | darkstar_ | <PROTECTED> |
| 05:09.04 | darkstar_ | <PROTECTED> |
| 05:09.04 | darkstar_ | UnicodeEncodeError: 'ascii' codec can't encode character u'\xf1' in position 584: ordinal not in range(128) |
| 05:09.06 | darkstar_ | same error |
| 05:10.21 | wgrant | What is type(content) at that point? |
| 05:11.48 | darkstar_ | it's difficult to know with omd_python |
| 05:11.58 | darkstar_ | I can't use it from the shell |
| 05:12.04 | darkstar_ | and I can't print it if I get an error |
| 05:12.06 | wgrant | req.write(str(type(content))) |
| 05:12.22 | darkstar_ | yes |
| 05:12.28 | darkstar_ | <PROTECTED> |
| 05:12.28 | darkstar_ | <PROTECTED> |
| 05:12.28 | darkstar_ | UnicodeEncodeError: 'ascii' codec can't encode character u'\xf1' in position 584: ordinal not in range(128) |
| 05:12.30 | darkstar_ | funny |
| 05:12.33 | darkstar_ | really funny |
| 05:12.42 | wgrant | ......... |
| 05:13.07 | wgrant | But... but..' |
| 05:13.18 | wgrant | There is no positon 584! |
| 05:13.24 | darkstar_ | exactly |
| 05:13.25 | darkstar_ | xD |
| 05:13.34 | wgrant | Oh. |
| 05:13.38 | wgrant | You are restarting Apache, right? |
| 05:14.00 | darkstar_ | look at this |
| 05:14.01 | wgrant | When something doesn't make sense, blame mod_python's cache. |
| 05:14.06 | darkstar_ | now works |
| 05:14.25 | darkstar_ | I'm restarting apache each minute |
| 05:14.30 | darkstar_ | removing sessions |
| 05:14.43 | darkstar_ | asking with vim |
| 05:17.02 | wgrant | Is the original thing working now? |
| 05:18.09 | darkstar_ | yes, but know it fails with same error in other place |
| 05:18.09 | darkstar_ | I've to reload apache... |
| 05:18.09 | darkstar_ | etc, etc, etc |
| 05:18.16 | wgrant | Yay, mod_python. |
| 05:18.43 | darkstar_ | Why is... |
| 05:18.48 | darkstar_ | as bad as is |
| 05:19.40 | wgrant | There is an option you can set that should mostly prevent it from caching. |
| 05:19.47 | darkstar_ | <PROTECTED> |
| 05:19.48 | darkstar_ | UnicodeDecodeError: 'ascii' codec can't decode byte 0xc3 in position 2: ordinal not in range(128) |
| 05:19.48 | wgrant | I don't remember what it is. |
| 05:20.05 | wgrant | That could be beause form[i] is already unicode. |
| 05:20.12 | wgrant | Er, wait. |
| 05:20.15 | wgrant | Already encoded. |
| 05:21.35 | darkstar_ | <PROTECTED> |
| 05:21.35 | darkstar_ | UnicodeDecodeError: 'ascii' codec can't decode byte 0xc3 in position 8: ordinal not in range(128) |
| 05:21.53 | darkstar_ | same error for form |
| 05:21.59 | wgrant | OK. |
| 05:22.01 | wgrant | Firstly. |
| 05:22.07 | wgrant | What's with all this a.__add__(b) |
| 05:22.10 | wgrant | Do you mean a + b? |
| 05:22.14 | darkstar_ | yes |
| 05:22.47 | wgrant | Why not use +? |
| 05:23.07 | darkstar_ | I've no reason... |
| 05:23.30 | darkstar_ | both do the same... |
| 05:23.59 | wgrant | Yes. But one is horrifyingly ugly. |
| 05:24.17 | darkstar_ | ok |
| 05:24.29 | darkstar_ | if you tell me how can I solve this errors |
| 05:24.38 | darkstar_ | I'll change it, I don't mind this |
| 05:24.47 | darkstar_ | add + ... |
| 05:25.30 | wgrant | So that is the real line on which the error occurs? mod_python isn't telling lies again? |
| 05:26.12 | darkstar_ | let me reload it... |
| 05:27.02 | darkstar_ | <PROTECTED> |
| 05:27.02 | darkstar_ | <PROTECTED> |
| 05:27.03 | darkstar_ | UnicodeDecodeError: 'ascii' codec can't decode byte 0xc3 in position 2: ordinal not in range(128) |
| 05:27.15 | darkstar_ | it fails with unicode() and without it |
| 05:27.45 | wgrant | What is session["formulario"]? |
| 05:28.00 | wgrant | Is session a mod_python session, and session["formulario"] just a dict? |
| 05:28.06 | darkstar_ | its a dictionary that contains few things |
| 05:28.15 | darkstar_ | sí |
| 05:28.20 | wgrant | Hmm. |
| 05:28.22 | darkstar_ | sorry |
| 05:28.22 | darkstar_ | yes |
| 05:28.25 | darkstar_ | it is |
| 05:28.28 | wgrant | I know that much :P |
| 05:28.47 | lifeless | darkstar_: don't cast to unicode |
| 05:28.58 | darkstar_ | without cast I get same error |
| 05:28.58 | lifeless | unicode(foo) never ever ever does the right thing |
| 05:29.03 | wgrant | Right, but if it fails without... |
| 05:29.22 | wgrant | It can't fail without. |
| 05:29.30 | lifeless | if its not unicode, you need an explicit decode, or to figure out why it wasn't unicode in the first place. |
| 05:29.38 | wgrant | So try again after removing the unicode(), and restarting Apache a couple of times. |
| 05:32.08 | darkstar_ | <PROTECTED> |
| 05:32.08 | darkstar_ | <PROTECTED> |
| 05:32.08 | darkstar_ | UnicodeDecodeError: 'ascii' codec can't decode byte 0xc3 in position 8: ordinal not in range(128) |
| 05:37.32 | darkstar_ | it will never works |
| 05:37.35 | darkstar_ | xD |
| 05:38.47 | wgrant | It will. |
| 05:38.51 | wgrant | I can't see how that is breaking. |
| 05:39.03 | wgrant | Unless session["formulario"] isn't actually a real dict. |
| 05:39.14 | darkstar_ | it is |
| 05:39.26 | darkstar_ | it only fails when tildes appears |
| 05:39.34 | wgrant | Right. |
| 05:40.06 | wgrant | Something on that line is trying to convert a unicode to a str. |
| 05:43.35 | darkstar_ | I'm trying and it's not possible |
| 05:43.46 | lifeless | " " is a str; combining a str and a unicode causes conversions |
| 05:44.08 | lifeless | u" " , a u": ", for starters |
| 05:44.12 | darkstar_ | >>> "hey ".__add__(" ".join( (unicode(u"hello"), "bye"))) |
| 05:44.12 | darkstar_ | u'hey hello bye' |
| 05:44.36 | darkstar_ | it's not neccesary as I can see in python |
| 05:44.46 | *** part/#storm lifeless (n=robertc@ppp245-86.static.internode.on.net) |
| 05:45.11 | wgrant | But that will be converting strs to unicodes. |
| 05:45.40 | darkstar_ | every str will be converted in unicode if an unicode adds to it |
| 05:45.47 | darkstar_ | as I've paste in an example |
| 05:45.49 | darkstar_ | and |
| 05:45.49 | wgrant | Right. |
| 05:45.55 | darkstar_ | <PROTECTED> |
| 05:45.56 | darkstar_ | <PROTECTED> |
| 05:45.56 | darkstar_ | UnicodeDecodeError: 'ascii' codec can't decode byte 0xc3 in position 17: ordinal not in range(128) |
| 05:46.02 | darkstar_ | doing it... same error |
| 05:46.40 | wgrant | Oh. Damn. It's a UnicodeDecodeError, not a UnicodeEncodeError. |
| 05:46.51 | darkstar_ | so... |
| 05:46.53 | wgrant | So something is trying to coerce a str to a unicode, not a unicode to a str. |
| 05:47.25 | wgrant | So the problem is that either i or session["formulario"][i] are encoded strs. |
| 05:47.41 | wgrant | They need to be decoded unicodes. |
| 05:47.49 | darkstar_ | so... |
| 05:48.02 | darkstar_ | I've to add decode.("utf8") |
| 05:48.04 | darkstar_ | ? |
| 05:51.22 | darkstar_ | oh my gooooooooooooooood |
| 05:51.25 | darkstar_ | now wooooooooooooorks |
| 05:51.41 | wgrant | Excellent! |
| 05:51.46 | darkstar_ | here you have a slave for ever |
| 05:51.47 | darkstar_ | hehe |
| 05:52.30 | darkstar_ | now I can go to sleep |
| 05:52.39 | darkstar_ | at 7:50 a.m |
| 05:52.51 | wgrant | Ouch! |
| 05:52.54 | darkstar_ | it's better later than never |
| 05:53.03 | darkstar_ | 7 hours later... |
| 05:53.06 | darkstar_ | hehehehe |
| 05:53.21 | darkstar_ | thanks you very very very much, wgrant |
| 05:53.27 | wgrant | darkstar_: No problem. |
| 05:53.38 | wgrant | darkstar_: You really want to read the Python unicode-related docs at some point, though. |
| 05:54.02 | darkstar_ | like we say in spain "I hope God thanks you with lot of childrens" |
| 05:54.30 | darkstar_ | I really want kill them right know |
| 05:54.45 | wgrant | Unicode is hard, unfortunately. |
| 05:54.50 | darkstar_ | I'm spending lot of time with mod_python |
| 05:54.58 | darkstar_ | and with codetypes |
| 05:54.59 | wgrant | (as are all non-ascii charsets) |
| 05:55.01 | wgrant | Yeah. |
| 05:55.03 | darkstar_ | and their respectives mothers |
| 05:55.05 | darkstar_ | hehehe |
| 05:55.19 | darkstar_ | the problem is that I really need tildes |
| 05:55.35 | darkstar_ | they are needed in spanish |
| 05:56.02 | darkstar_ | á, é, í, ó, ú, and ñ |
| 05:56.06 | wgrant | Of course. |
| 05:56.10 | darkstar_ | and ü |
| 05:56.13 | wgrant | Applications should support Unicode. |
| 05:56.24 | wgrant | It's just difficult to work it all out the first time. |
| 05:56.31 | wgrant | Eventually you get used to dealing with Unicode everywhere. |
| 05:57.19 | darkstar_ | now I can post and edit articles |
| 05:57.22 | darkstar_ | very good |
| 05:58.01 | darkstar_ | I'll finish this page in 2012 if I'm lucky |
| 09:05.34 | *** join/#storm wgrant (n=wgrant@ubuntu/member/wgrant) |
| 09:26.06 | *** join/#storm stub (n=stub@ppp-58-8-4-146.revip2.asianet.co.th) |
| 10:51.45 | *** join/#storm andrea-bs (n=andrea@ubuntu/member/beeseek.developer.andrea-bs) |
| 13:30.54 | *** join/#storm keppla (n=keppla@i577B35DB.versanet.de) |
| 13:33.09 | keppla | x |
| 13:33.13 | keppla | s/x/y |
| 17:39.21 | *** join/#storm keppla (n=keppla@i577B35DB.versanet.de) |
| 22:08.05 | *** join/#storm mup (n=mup@li37-90.members.linode.com) |
| 23:47.29 | *** join/#storm darkstar_ (n=nobody@198.Red-213-98-22.staticIP.rima-tde.net) |
| 23:47.33 | darkstar_ | hi all |
| 23:47.51 | darkstar_ | wgrant, I solved my problems, but now I've another joining 2 tables |
| 23:48.12 | darkstar_ | I need info from a table that is a relationship between 2 tables |
| 23:48.28 | darkstar_ | and something happends, I get a 2 tuple result with no data |
| 23:48.48 | darkstar_ | I've been looking for how to do it in storm webpage, but it doesn't work |
| 23:49.54 | *** join/#storm keppla1 (n=keppla@i577B00B6.versanet.de) |