mamweb/seminar/ovvpfile.py

# -*- coding: utf-8 -*-

class OvvpFile(object):
    def __init__(self):
        self.headers = {}
        self.columns = []
        self.rows = []

    def to_lines(self):
        # header
        for hk in sorted(self.headers.keys()):
            yield '%s\t%s\n' % (hk, self.headers[hk])
        yield '\n'
        # columns
        yield '\t'.join([c for c in self.columns]) + '\n'
        # rows
        for r in self.rows:
            yield '\t'.join([r[c] for c in self.columns]) + '\n'

    def to_string(self):
        return ''.join(self.to_lines())

    def parse_from(self, source, with_headers=True):
        "Parse data from file, string or line iterator, overwriting self"
        if isinstance(source, str) or isinstance(source, unicode):
            return self.parse_from(source.split('\n'))
            
        it = iter(source)

        # header
        self.headers = {}
        if with_headers:
            for r in it:
                if isinstance(r, str):
                    r = r.decode('utf8')
                assert isinstance(r, unicode)
                r = r.rstrip('\n')
                if r == u"":
                    break
                k, v = r.split(u'\t', 1)
                self.headers[k] = v

        # columns
        r = it.next()
        if isinstance(r, str):
            r = r.decode('utf8')
        self.columns = [cn.strip() for cn in r.split(u'\t') if cn.strip() != ""]

        # rows
        self.rows = []
        for r in it:
            if isinstance(r, str):
                r = r.decode('utf8')
            r = r.rstrip('\n')
            if not r:
                break
            rtup = r.split(u'\t')
            rdict = {}
            for ci in range(len(self.columns)):
                rdict[self.columns[ci]] = rtup[ci]
            self.rows.append(rdict)
        


def parse(source, with_headers=True):
    o = OvvpFile()
    o.parse_from(source, with_headers=with_headers)
    return o