Re: gEDA-user: gEDA file format - couple of questions on attributes.
Kai-Martin Knaak k...@lilalaser.de writes: Rather than restrict the input to a smallish subset, the code should deal gracefully with as much of UTF8 as possible. Um, the code deals just fine with attribute names containing arbitrary UTF-8. It's just not *recommended* to use attribute names that don't match the regexp I supplied. What is the rationale to this restriction? It's not a restriction, it's a recommendation. The main reasons are that: 1) We need to support gnetlist backends when gEDA is compiled against Guile 1.8.x, which doesn't understand UTF-8. BTW, if you want to use non-ASCII characters in your attributes, I highly recommend to use Guile 2.x. 2) In the future, we may wish to do clever things with attribute names, perhaps giving characters like '#', '/' or ':' on the left hand side of the '=' in an attribute special meaning to gnetlist. If you stick to the regexp I provided, you will be future-proof. To repeat: you can use any Unicode codepoints you like in attribute names at the moment. However, at the moment it's safest to stick to '^[a-z][a-z0-9_-\.]*$'. Regards, Peter P.S. To clarify, attributes *must* match '^(.*[^ ])=([^ ].*)$' (multiline) at the moment. -- Peter Brett pe...@peter-b.co.uk Remote Sensing Research Group Surrey Space Centre ___ geda-user mailing list geda-user@moria.seul.org http://www.seul.org/cgi-bin/mailman/listinfo/geda-user
Re: gEDA-user: gEDA file format - couple of questions on attributes.
On Jun 28, 2011, at 4:17 AM, Peter Brett wrote: 1) We need to support gnetlist backends when gEDA is compiled against Guile 1.8.x, which doesn't understand UTF-8. BTW, if you want to use non-ASCII characters in your attributes, I highly recommend to use Guile 2.x. Actually, UTF-8 attributes (names and values) work fine with Guile 1.8-based gEDA. The clever design of UTF-8 makes code intended for ASCII just work so long as it doesn't need to pick wide characters out of a string. John Doty Noqsi Aerospace, Ltd. http://www.noqsi.com/ j...@noqsi.com ___ geda-user mailing list geda-user@moria.seul.org http://www.seul.org/cgi-bin/mailman/listinfo/geda-user
Re: gEDA-user: gEDA file format - couple of questions on attributes.
This questions intereresting for me too, so it would be nice if somebody of Guru could comment them :) Hi there, I'm still intermittently working on a parser in Python for the gEDA/gaf file format and I have a couple of questions that I hope the list can answer. 1) are there any restrictions on the characters in attribute names and values (other than no = in the name) ? As I see in gschem sources slot attribute have some special meaning. But seems it just providing some special processing for sloted objects (couldnt't imagine what is this). Also check here: [1]http://geda.seul.org/wiki/geda:master_attributes_list [2]http://geda.seul.org/wiki/geda:na_howto 2) Can attributes be multiline text or are they always single line text (but perhaps several)? As I see attributes are always one line. Attribute editor is one line length. In the code I don't see any special processing for multiline attributes. So seems 1 attribute = 1 line. Regards, Pavlo References 1. http://geda.seul.org/wiki/geda:master_attributes_list 2. http://geda.seul.org/wiki/geda:na_howto ___ geda-user mailing list geda-user@moria.seul.org http://www.seul.org/cgi-bin/mailman/listinfo/geda-user
Re: gEDA-user: gEDA file format - couple of questions on attributes.
[Re-sending to list, since there's been additional interest.] Gareth Edwards gar...@edwardsfamily.org.uk writes: Hi there, I'm still intermittently working on a parser in Python for the gEDA/gaf file format and I have a couple of questions that I hope the list can answer. 1) are there any restrictions on the characters in attribute names and values (other than no = in the name) ? At the moment, gEDA allows names to include any UTF-8 character other than '=' or \0, but (for some reason) they must not end with a space. However, I strongly recommend that you only use attribute names that match: ^[a-z][a-z0-9_-\.]*$ Since you're writing a parser, consider emitting a warning if you encounter an attribute name which doesn't. 2) Can attributes be multiline text or are they always single line text (but perhaps several)? Yes, they can be multiline (any UTF-8 character other than \0). Also, they (randomly) may not start with a space. Since the gEDA file format requires UTF-8, please make sure your parser raises the alarm if it encounters encoding errors. Cheers, Peter P.S. I'm hoping to tighten up the rules on attributes soon-ish. -- Peter Brett pe...@peter-b.co.uk Remote Sensing Research Group Surrey Space Centre ___ geda-user mailing list geda-user@moria.seul.org http://www.seul.org/cgi-bin/mailman/listinfo/geda-user
Re: gEDA-user: gEDA file format - couple of questions on attributes.
1) are there any restrictions on the characters in attribute names and values (other than no = in the name) ? The implementation of the attribute name, value parser is in the function around line 357 of ./libgeda/src/o_attrib.c file: 357 gboolean 358 o_attrib_string_get_name_value (const gchar *string, gchar **name_ptr, gchar **value_ptr) The above parser function, though not perfect, allows flexibility within UTF-8 encoded string. As such, within the T object string portion, the first detected = with no space char on either adjacent side of it, separates the attribute name and the attribute value. attribute value portion can contain any UTF-8 encoded string (except an immediate '\0'), including =. Other than the above, there seems to be no other restrictions imposed by the frontend. Beyond the gschem file format, there could be other restriction on attributes imposed by some processing entity (DRC, etc) and backend netlister. But those are backend specific, and upto the particular backend specific needs. 2) Can attributes be multiline text or are they always single line text (but perhaps several)? Attribute or text can be multiline as stated in the text section of: http://geda.seul.org/wiki/geda:file_format_spec The num_lines field in the T object type line indicates the number of lines in the string. ___ geda-user mailing list geda-user@moria.seul.org http://www.seul.org/cgi-bin/mailman/listinfo/geda-user
Re: gEDA-user: gEDA file format - couple of questions on attributes.
Peter Brett wrote: However, I strongly recommend that you only use attribute names that match: ^[a-z][a-z0-9_-\.]*$ Ouch. Every string in the user interface should potentially be subject to localization. Every input field should accept local characters. Since attributes can be introduced by th user, the add attribute function is effectively an input field. Confinement to ascii chars prevents proper spelling of non-english words. What is the rationale to this restriction? IIRC, the glibc string operations are TTF8 aware. Rather than restrict the input to a smallish subset, the code should deal gracefully with as much of UTF8 as possible. ---)kaimartin(--- -- Kai-Martin Knaak Email: k...@familieknaak.de Öffentlicher PGP-Schlüssel: http://pool.sks-keyservers.net:11371/pks/lookup?search=0x6C0B9F53 ___ geda-user mailing list geda-user@moria.seul.org http://www.seul.org/cgi-bin/mailman/listinfo/geda-user
Re: gEDA-user: gEDA file format - couple of questions on attributes.
However, I strongly recommend that you only use attribute names that match: Confinement to ascii chars prevents proper spelling of non-english words. What is the rationale to this restriction? IIRC, the glibc string operations are TTF8 aware. Rather than restrict the input to a smallish subset, the code should deal gracefully with as much of UTF8 as possible. We're not talking about attribute values or plain text strings. We're talking about attribute *names* - things that other bits of software need to use as keys for understanding what the values are for. Since they're mostly for other programs to read, a bit more restriction is appropriate. ___ geda-user mailing list geda-user@moria.seul.org http://www.seul.org/cgi-bin/mailman/listinfo/geda-user
