[parse_csv]
| Link | parse_csv |
| Author | Johan Solve |
| Category | File |
| Version | 8.x |
| License | Public Domain |
| Posted | 25 Mar 2012 |
| Updated | 25 Mar 2012 |
| More by this author... | |
Description
Custom tag to parse text files with any field delimiters, for example tab, comma (the default), semicolon etc. It uses a state machine to be able to handle optional field delimiters, for example when a CSV file only has quotes around field values when a comma is present in the data. Outputs either array of arrays or array of maps. Note: to output array of maps, the first row of data must contain field names.
Sample Usage
parse_csv(file_read(filename.csv), -linebreak='\r\n', -fieldseparator=',', -fielddelimiter='"', -escapecharacter='\\', -firstrowfieldnames, -rowformat='map', -optionalfielddelimiters);
Source Code
Click the "Download" button below to retrieve a copy of this tag, including the complete documentation and sample usage shown on this page. Place the downloaded ".inc" file in your LassoStartup folder, restart Lasso, and you can begin using this tag immediately.
define_tag('parse_csv', -required='input'
, -optional='linebreak', -type='string', -copy
, -optional='fielddelimiter', -type='string', -copy
, -optional='fieldseparator', -type='string', -copy
, -optional='escapecharacter', -type='string', -copy
, -optional='optionaldelimiters', -copy
, -optional='firstrowfieldnames', -copy
, -optional='rowformat', -type='string', -copy
);
local( 'state'='out'
, 'output'=array
, 'row'=array
, 'field'=string
, 'char'=string
, 'escaped'=false
, 'unquoted'=false
, 'fieldnames'=array
);
!local_defined('linebreak') ? local('linebreak'='\n');
!local_defined('fielddelimiter') ? local('fielddelimiter'='"');
!local_defined('fieldseparator') ? local('fieldseparator'=',');
!local_defined('escapecharacter') ? local('escapecharacter'='\\');
!local_defined('rowformat') ? local('rowformat'='array');
local('optionaldelimiters'=(local_defined('optionaldelimiters') && #optionaldelimiters != false));
local('firstrowfieldnames'=(local_defined('firstrowfieldnames') && #firstrowfieldnames != false));
iterate(#input -> split(''), #char);
select(#state);
case('out');
if(#char==#fielddelimiter);
#state='in';
else(#char==#fieldseparator);
if(#firstrowfieldnames && #rowformat=='map' && #fieldnames -> size);
if(#fieldnames -> size >= #row -> size + 1);
#row -> insert(#fieldnames -> get(#row -> size + 1) = #field);
/if;
else;
#row -> insert(#field);
/if;
#field=string;
else(#char==#linebreak);
if(#firstrowfieldnames && #rowformat=='map' && #fieldnames -> size);
if(#fieldnames -> size >= #row -> size + 1);
#row -> insert(#fieldnames -> get(#row -> size + 1) = #field);
/if;
else;
#row -> insert(#field);
/if;
#output -> insert(#row);
#field=string;
#rowformat == 'map' ? #row = map | #row = array;
else(#optionaldelimiters);
#state='in';
#unquoted=true;
#field += #char;
/if;
case('in');
if(!#escaped && #char == #escapecharacter);
#escaped=true;
else(#char==#fielddelimiter && !#escaped && !#unquoted);
#state='out';
else(#char==#fieldseparator && #optionaldelimiters && #unquoted && !#escaped);
#state='out';
#unquoted=false;
if(#firstrowfieldnames && #rowformat=='map' && #fieldnames -> size);
if(#fieldnames -> size >= #row -> size + 1);
#row -> insert(#fieldnames -> get(#row -> size + 1) = #field);
/if;
else;
#row -> insert(#field);
/if;
#field=string;
else(#char==#linebreak && #optionaldelimiters && #unquoted && !#escaped);
#state='out';
#unquoted=false;
if(#firstrowfieldnames && #rowformat=='map' && #fieldnames -> size);
if(#fieldnames -> size >= #row -> size + 1);
#row -> insert(#fieldnames -> get(#row -> size + 1) = #field);
/if;
else;
#row -> insert(#field);
/if;
#output -> insert(#row);
#field=string;
#rowformat == 'map' ? #row = map | #row = array;
#field=string;
else;
#field += #char;
#escaped=false;
/if;
/select;
if(#firstrowfieldnames && #output -> size == 1 && #fieldnames -> size == 0);
#fieldnames = #output -> get(1);
#output -> remove(1);
else(!#firstrowfieldnames);
#rowformat='array';
/if;
/iterate;
if(#field -> size);
if(#firstrowfieldnames && #rowformat=='map' && #fieldnames -> size);
if(#fieldnames -> size >= #row -> size + 1);
#row -> insert(#fieldnames -> get(#row -> size + 1) = #field);
/if;
else;
#row -> insert(#field);
/if;
/if;
#row -> size ? #output -> insert(#row);
return(#output);
/define_tag;
Related Tags
Comments
No comments
