Syntax Parameters

Syntax contains parameters that are used in reading or writing edi-files.
The complete list of syntax parameters including default values is in bots/grammar.py (in the classes of the editypes).
Syntax is a python dict (dictionary).

Example of syntax parameters:

syntax = {
        'charset'                  : 'utf-8', #character set is utf-8
        'checkfixedrecordtooshort' : True,    #check if fixed record is to short
        'indented'                 : True,    #xml: produced indented output
        'decimaal'                 : ',',     #decimal sign is ', '
        }

Usage and Overriding

Syntax parameters can be set at different places; these settings override (somewhat like CSS). The order in which overriding is done:

default values are in bots/grammars.py (per editype). Do not change these values here.
envelope grammar (eg for x12: bots/usersys/grammars/x12/x12.py, for edifact: bots/usersys/grammars/edifact/edifact.py)
message grammar
frompartner grammar (eg in bots/usersys/partners/x12/partnerID.py)
topartner grammar (eg in bots/usersys/partners/x12/partnerID.py)

Example 1: edifact charset

default value is UNOA
value in envelope (edifact.py) is UNOA
for invoices: a description is used so the message grammar for invoices has charset UNOC
retailer ABC insists on receiving invoices as UNOA, so this is indicated in the topartner grammar.

Example 2: x12 element separator

in grammar.py: 'field_sep':'*' (bots default value)
in x12.py: 'field_sep':'|' (default value company uses when sending x12)
retailer ABC insists: 'field_sep':'\x07' (that is \a, or BEL)

List of most useful Syntax Parameters

Parameter	Direction	Description
add_crlfafterrecord_sep	Out	Put an extra character after a record/segment separator. Value: string, typically `\n` or `\r\n`. I use this for x12/edifact while testing: output is easier to read. Most partners can handles these file, but they are slightly bigger. See also parameter `indented`.
acceptspaceinnumfield	In	Do not raise error when numeric field contains only spaces but assume value is 0.
allow_lastrecordnotclosedproperly	In	(csv) Allows last record not to have record separator.
charset	In	Charset to use; (edifact, xml) is overridden by charset-declaration in content.
	Out	Charset to use in output. Bots is quite strict in this.
checkcharsetin	In	What to do for chars not in charset. Possible values ‘strict’ (gives error) or ‘ignore’ (skip the characters not in the charset) or ‘botsreplace’ (replace with char as set in bots.ini; default is space)
checkcharsetout	Out	What to do for chars not in charset. Possible values ‘strict’ (gives error) or ‘ignore’ (skip the characters not in the charset) or ‘botsreplace’ (replace with char as set in bots.ini; default is space)
checkfixedrecordtoolong	In	(fixed) Warn if record is too long. Possible values: True/False, default: True.
checkfixedrecordtooshort	In	(fixed) Warn if record is too short. Possible values: True/False, default: False.
checkunknownentities	In/Out	(xml, json) Skip unknown attributes/elements (instead of raising an error).
contenttype	Out	Content-type of translated file; used as mime-envelope of email.
decimaal	In/Out	Decimal point; default is ‘.’. For edifact: read from UNA-string if present.
endrecordID	In	(fixed) End position of record ID. Possible value: number, default 3. See startrecordID.
envelope	Out	Envelope to use. If nothing specified: no envelope - files are just copied/appended. For csv output, use the value ‘csvheader’ to include a header line with field names.
envelope-template	Out	(template/html) The template for the envelope.
escape	In/Out	Escape character used. Default: edifact: ‘?’.
field_sep	In/Out	Field separator. Default: edifact: ‘+’; csv: ‘:’ x12: ‘*’; tradacoms: ‘+’.
forcequote	Out	(csv) Possible values: 1 (quote only if necessary); 1 (always quote), 2 (quote only alfanum). Default is 1.
forceUNA	Out	(edifact) Always use UNA-segment in header, even if not needed. Possible values: True, False. Default: False.
indented	Out	(xml, json) Indent message for human readability. Nice while testing. Indented messages are syntactically OK, but are much bigger. Possible values: True, False. Default: False. See also add_crlfafterrecord_sep.
merge	Out	If merge is True: merge translated messages to one file (for same sender, receiver, messagetype, channel, etc). Related: envelope. Possible values: True, False.
namespace_prefixes	Out	(xml) To over-ride default namespace prefixes (ns0, ns1, etc.) for outgoing xml. It is a list, consisting of tuples, each tuple consists of prefix and uri. Example: `'namespace_prefixes':[('orders','http://www.company.com/EDIOrders'),]`
noBOTSID	In/Out	(csv) Use if records contain no real record ID.
output	Out	(template) Possible value: ‘xhtml-strict’.
pass_all	In	(csv, fixed) If only one recordtype and no nextmessageblock; False: pass record for record to mapping; True: Pass all records to mapping.
preprocess_lex	In	Lex preprocessing via user exit with a callable. Default: False.
preprocess_nodes	In	Node preprocessing via user exit with a callable. Default: False.
quote_char	In/Out	(csv) Character used as quote symbol. Default: ‘’’ (single quote).
record_sep	In/Out	Character used as record separator. Defaults: edifact: ‘’’ (single quote); fixed: ‘n’; x12: ‘~’.
record_tag_sep	Out	(tradacoms) Separator used after segment tag. Defaults: ‘=’.
replacechar	Out	(x12) If a separator value is found in the data, replace with this character. Default: ‘’ (raise error when separator value in data).
skip_char	In	Character(s) to skip, not interpreted when reading file. Default in edifact, x12 and tradacoms: ‘rn’
skip_firstline	In	(csv) Skip first line (often contains field names). Possible values: True/False/Integer number of lines to skip, default: False. True skips 1 line.
startrecordID	In	(fixed) Start position of record ID; Possible value: number, default 0. See endrecordID.
strict_syntax_check	In	For files with variable lenght records, except CSV: don’t allow whitespaces between records, double record separators or alphanumeric repeating separator. Possible values: True/False. Default: False.
strip_value	In	(csv, edifact, x12, tradacom) Possible values: True/False. True removes any leading, and trailing whitespaces from the record. Default: False.
template_engine	Out	(template) Template to use for HTML-output. Possible values: django or genshi. Default: genshi.
triad	In	Triad (thousands) symbol used (e.g. ‘1,000,048.35’). If specified, this symbol is skipped when parsing numbers. By default numbers are expected to come without thousands separators.
version	Out	(edifact, x12) Version of standard generate. Value: string, typically: ‘3’ in edifact or ‘004010’ for x12.
wrap_length	Out	Wraps the output to a new line when it exceeds this length. Possible value: number, default 0. Typically used in conjunction with ‘add_crlfafterrecord_sep’:’’ (blank). Note: does not affect envelope of message.