Experiments in reading CSV documents.

Clone the repo and run make . That will build and run the tests.

Shrdlu breaks down the problem of parsing CSV data into three stages.

1. A field parser, which parses single fields of a CSV document.

2. Using the Field parser to make a record parser, which parses one CSV record.

3. Using the record parser to make a document parser, which reads an entire CSV document.

Shrdlu is easy to use. Just do extractions from an input stream into one of the parser components, like so:

// Read one field.
Field_parser fp;
input_stream >> fp;
// now do something with fp or with fp.get_field().

// Read one record.
Record_parser rp;
input_sream >> rp;
// do something with rp or with rp.get_record();

// Read one document.
Document_parser dp;
input_stream >> dp;
// do something with dp or with dp.get_document();

Look in test.cc for details.