diff options
author | Yves Fischer <yvesf-git@xapek.org> | 2016-07-23 18:42:06 +0200 |
---|---|---|
committer | Yves Fischer <yvesf-git@xapek.org> | 2016-07-23 18:42:40 +0200 |
commit | 76289024543fbfb6df3d07f61f15eabc623dd246 (patch) | |
tree | 5f7e7a86e42f26e8afe3b4437f77e9e4100280ee /pyinflux | |
parent | f019accc6f20ef97e95b9b6a3f776aefb5ebd62b (diff) | |
download | pyinflux-76289024543fbfb6df3d07f61f15eabc623dd246.tar.gz pyinflux-76289024543fbfb6df3d07f61f15eabc623dd246.zip |
restructure code
rename module to pyinflux
split in client and parser package. the parser depends also on funcparserlib
contains 3 examples of strange influxdb behavior
Diffstat (limited to 'pyinflux')
-rw-r--r-- | pyinflux/__init__.py | 1 | ||||
-rw-r--r-- | pyinflux/client/__init__.py | 149 | ||||
-rw-r--r-- | pyinflux/parser/__init__.py | 222 |
3 files changed, 372 insertions, 0 deletions
diff --git a/pyinflux/__init__.py b/pyinflux/__init__.py new file mode 100644 index 0000000..bc63beb --- /dev/null +++ b/pyinflux/__init__.py @@ -0,0 +1 @@ +# empty
\ No newline at end of file diff --git a/pyinflux/client/__init__.py b/pyinflux/client/__init__.py new file mode 100644 index 0000000..b485861 --- /dev/null +++ b/pyinflux/client/__init__.py @@ -0,0 +1,149 @@ +import typing +import io +import re +from urllib.request import urlopen +from urllib.parse import quote as urlquote, urlencode +import json +import codecs + + +class Line(object): + def __init__(self, key, tags, fields, timestamp=None): + self.key = key + self.tags = tags + self.fields = fields + self.timestamp = timestamp + + if isinstance(self.tags, dict): + self.tags = self.tags.items() + + if isinstance(self.fields, dict): + self.fields = self.fields.items() + + @staticmethod + def escape_identifier(string): + return re.sub(r'([\\,= ])', '\\\\\\1', string) + + @staticmethod + def escape_tags(taglist): + return ",".join(map(lambda kv: + (Line.escape_identifier(kv[0]) + "=" + Line.escape_identifier(kv[1])), + taglist)) + + @staticmethod + def escape_value(obj): + if (isinstance(obj, float) or + isinstance(obj, int) or + isinstance(obj, bool)): + return str(obj) + else: + obj = str(obj) + return "\"" + obj.replace("\\", "\\\\").replace("\"", "\\\"") + "\"" + + @staticmethod + def escape_fields(kvlist): + def escape_key(string): + return re.sub(r'(["\\,= ])', '\\\\\\1', string) + + return ",".join( + map(lambda kv: escape_key(kv[0]) + "=" + Line.escape_value(kv[1]), + kvlist)) + + def __repr__(self): + """ + >>> print(repr(Line('test', [('a','b')], [('value','asd\\\\')]))) + <Line key=test tags=[('a', 'b')] fields=[('value', 'asd\\\\')] timestamp=None> + """ + return "<{} key={} tags={} fields={} timestamp={}>".format( + self.__class__.__name__, self.key, self.tags, self.fields, self.timestamp) + + def __str__(self): + """ + >>> print(Line('test', [('a','b')], [('value','asd\\\\')])) + test,a=b value="asd\\\\" + """ + result = self.escape_identifier(self.key) + + if self.tags: + result += "," + result += self.escape_tags(self.tags) + + if self.fields: + result += " " + result += self.escape_fields(self.fields) + + if self.timestamp: + result += " " + result += str(self.timestamp) + + return result + + +class QueryResultOption: + CODEC = codecs.getreader('utf-8') + + def __init__(self, exec_func: typing.Callable[[], io.IOBase]): + self.exec_func = exec_func + self._json = None + self._text = None + + def as_json(self): + if self._json is None: + fh = self.CODEC(self.exec_func()) + self._json = json.load(fh) + fh.close() + return self._json + + def as_text(self): + if self._text is None: + fh = self.CODEC(self.exec_func()) + self._text = fh.read() + fh.close() + return self._text + + +class Influx: + def __init__(self, host: str, port: int = 8086, username: str = None, password: str = None): + """ + :param username: username and password: + :param password: if set both must be set + """ + self._write_url = "http://{host}:{port}/write?".format(**locals()) + self._query_url_get = "http://{host}:{port}/query?".format(**locals()) + self._query_url_post = "http://{host}:{port}/query".format(**locals()) + if username and password: + self._write_url += 'username=' + username + '&password' + password + '&' + self._query_url_get += 'username=' + username + '&password' + password + '&' + self._query_url_post += '?username=' + username + '&password' + password + + def write_db(self, db: str, lines: [Line]): + url = self._write_url + "db=" + urlquote(db) + request_data = "\n".join(map(str, lines)).encode('utf-8') + with urlopen(url, request_data) as fh: + response = fh.read() + return response.decode('utf-8') + + def query_db(self, db: str, query: str) -> QueryResultOption: + def get_fh() -> io.IOBase: + url = self._query_url_get + 'db=' + urlquote(db) + '&q=' + urlquote(query) + return urlopen(url) + + return QueryResultOption(get_fh) + + def execute(self, query: str) -> QueryResultOption: + def get_fh() -> io.IOBase: + return urlopen(self._query_url_post, urlencode({'q': query}).encode('utf-8')) + + return QueryResultOption(get_fh) + + +class InfluxDB(Influx): + def __init__(self, db: str, host: str, port: int = 8086, username: str = None, password: str = None): + super().__init__(host, port, username, password) + self._db = db + + def write(self, lines: [Line]): + return self.write_db(self._db, lines) + + def query(self, query: str): + return self.query_db(self._db, query) diff --git a/pyinflux/parser/__init__.py b/pyinflux/parser/__init__.py new file mode 100644 index 0000000..2de6c0e --- /dev/null +++ b/pyinflux/parser/__init__.py @@ -0,0 +1,222 @@ +from functools import reduce + +try: + from funcparserlib.lexer import make_tokenizer + from funcparserlib.parser import ( + some, maybe, many, finished, skip, NoParseError) +except ImportError as e: + print("Missing funcparserlib library. You need to install the 'parser' extra dependency set.") + raise e + +from pyinflux import client + + +def parse_lines(lines): + """ + Parse multiple Write objects separeted by new-line character. + + >>> print(LineParser.parse("foo b=1")) + foo b=1 + + >>> lines = [] + >>> lines += ['cpu field=123'] + >>> lines += ['cpu,host=serverA,region=us-west field1=1,field2=2'] + >>> lines += ['cpu,host=serverA,region=us-west field1=1,field2=2 1234'] + >>> print("\\n".join(map(str, parse_lines("\\n".join(lines))))) + cpu field=123 + cpu,host=serverA,region=us-west field1=1,field2=2 + cpu,host=serverA,region=us-west field1=1,field2=2 1234 + """ + writes = map(LineParser.parse, lines.split("\n")) + return list(writes) + + +class LineParser(object): + specs = [ + ('Comma', (r',',)), + ('Space', (r' ',)), + ('Equal', (r'=',)), + ('Quote', (r'"',)), + ('Escape', (r'\\',)), + ('Int', (r'[0-9]+(?![0-9\.])',)), + ('Float', (r'-?(\.[0-9]+)|([0-9]+(\.[0-9]*)?)',)), + ('Char', (r'.',)), + ] + + @classmethod + def tokenize(klass, line : str): + tokenizer = make_tokenizer(klass.specs) + return list(tokenizer(line)) + + @staticmethod + def parse(line): + """ + Parse a line from the POST request into a Write object. + + >>> line='cpu a=1'; LineParser.parse(line); print(LineParser.parse(line)) + <Line key=cpu tags=[] fields=[('a', 1)] timestamp=None> + cpu a=1 + + >>> print(LineParser.parse('yahoo.CHFGBP\\=X.ask,tag=foobar value=10.2')) + yahoo.CHFGBP\=X.ask,tag=foobar value=10.2 + + >>> LineParser.parse('cpu,host=serverA,region=us-west foo="bar"') + <Line key=cpu tags=[('host', 'serverA'), ('region', 'us-west')] fields=[('foo', 'bar')] timestamp=None> + + >>> print(LineParser.parse('cpu host="serverA",region="us-west"')) + cpu host="serverA",region="us-west" + + >>> line='cpu\\,01 host="serverA",region="us-west"'; \\ + ... LineParser.parse(line); print(LineParser.parse(line)) + <Line key=cpu,01 tags=[] fields=[('host', 'serverA'), ('region', 'us-west')] timestamp=None> + cpu\,01 host="serverA",region="us-west" + + >>> LineParser.parse('cpu host="server A",region="us west"') + <Line key=cpu tags=[] fields=[('host', 'server A'), ('region', 'us west')] timestamp=None> + + >>> line='cpu ho\\=st="server A",region="us west"'; \\ + ... LineParser.parse(line); print(LineParser.parse(line)) + <Line key=cpu tags=[] fields=[('ho=st', 'server A'), ('region', 'us west')] timestamp=None> + cpu ho\=st="server A",region="us west" + + >>> print(LineParser.parse('cpu,ho\=st=server\ A field=123')) + cpu,ho\=st=server\ A field=123 + + # error: double name is accepted + >>> print(LineParser.parse('cpu,foo=bar,foo=bar field=123,field=123')) + cpu,foo=bar,foo=bar field=123,field=123 + + >>> print(LineParser.parse('cpu field12=12')) + cpu field12=12 + + >>> print(LineParser.parse('cpu field12=12 123123123')) + cpu field12=12 123123123 + + >>> try: print(LineParser.parse('cpu field12=12 1231abcdef123')) + ... except NoParseError: pass + + >>> print(LineParser.parse('cpu,x=3,y=4,z=6 field\ name="HH \\\\\\"World",x="asdf foo"')) + cpu,x=3,y=4,z=6 field\\ name="HH \\"World",x="asdf foo" + + >>> print(LineParser.parse("cpu,x=3 field\ name=\\"HH \\\\\\"World\\",x=\\"asdf foo\\"")) + cpu,x=3 field\\ name="HH \\"World",x="asdf foo" + + >>> print(LineParser.parse("cpu foo=\\"bar\\" 12345")) + cpu foo="bar" 12345 + + >>> line='"measurement\ with\ quotes",tag\ key\ with\ spaces=tag\,value\,with"commas" field_key\\\\\\="string field value, only \\\\" need be quoted"'; \\ + ... LineParser.parse(line); print(LineParser.parse(line)) + <Line key="measurement with quotes" tags=[('tag key with spaces', 'tag,value,with"commas"')] fields=[('field_key\\\\', 'string field value, only " need be quoted')] timestamp=None> + "measurement\ with\ quotes",tag\ key\ with\ spaces=tag\,value\,with"commas" field_key\\\\="string field value, only \\\" need be quoted" + + >>> LineParser.parse('disk_free value=442221834240,working\ directories="C:\My Documents\Stuff for examples,C:\My Documents"') + <Line key=disk_free tags=[] fields=[('value', 442221834240), ('working directories', 'C:\\\\My Documents\\\\Stuff for examples,C:\\\\My Documents')] timestamp=None> + + >>> LineParser.parse('disk_free value=442221834240,working\ directories="C:\My Documents\Stuff for examples,C:\My Documents" 123') + <Line key=disk_free tags=[] fields=[('value', 442221834240), ('working directories', 'C:\\\\My Documents\\\\Stuff for examples,C:\\\\My Documents')] timestamp=123> + + >>> print(LineParser.parse('foo,foo=2 "field key with space"="string field"')) + foo,foo=2 field\ key\ with\ space="string field" + + >>> print(LineParser.parse('foo,foo=2 field_key\\\\\\="string field"')) + foo,foo=2 field_key\\\\="string field" + + >>> print(LineParser.parse('foo,foo=2 field_key="string\\\\" field"')) + foo,foo=2 field_key="string\\" field" + + >>> line='foo field0="tag",field1=t,field2=true,field3=True,field4=TRUE'; \\ + ... LineParser.parse(line); print(LineParser.parse(line)) + <Line key=foo tags=[] fields=[('field0', 'tag'), ('field1', True), ('field2', True), ('field3', True), ('field4', True)] timestamp=None> + foo field0="tag",field1=True,field2=True,field3=True,field4=True + + >>> line='foo field1=f,field2=false,field3=False,field4=FALSE,field5="fag"'; \\ + ... LineParser.parse(line); print(LineParser.parse(line)) + <Line key=foo tags=[] fields=[('field1', False), ('field2', False), ('field3', False), ('field4', False), ('field5', 'fag')] timestamp=None> + foo field1=False,field2=False,field3=False,field4=False,field5="fag" + + >>> line='"measurement\ with\ quotes",tag\ key\ with\ spaces=tag\,value\,with"commas" field_key\\\\\\="string field value, only \\\\" need be quoted"'; \\ + ... LineParser.parse(line); print(LineParser.parse(line)) + <Line key="measurement with quotes" tags=[('tag key with spaces', 'tag,value,with"commas"')] fields=[('field_key\\\\', 'string field value, only " need be quoted')] timestamp=None> + "measurement\ with\ quotes",tag\ key\ with\ spaces=tag\,value\,with"commas" field_key\\\\="string field value, only \\" need be quoted" + + >>> LineParser.parse('"measurement\ with\ quotes" foo=1') + <Line key="measurement with quotes" tags=[] fields=[('foo', 1)] timestamp=None> + + >>> print(LineParser.parse('K.5S,Ccpvo="eSLyE" value="7F\\\\\\\\\\\\""')) + K.5S,Ccpvo="eSLyE" value="7F\\\\\\\\\\\"" + + >>> print(LineParser.parse('K.5S,Ccpvo=a\\ b value=1')) + K.5S,Ccpvo=a\\ b value=1 + """ + + tokval = lambda t: t.value + joinval = "".join + someToken = lambda type: some(lambda t: t.type == type) + someCharValue = lambda string: \ + reduce(lambda a, b: a + b, + map(lambda char: + some(lambda t: t.value == char) >> tokval, + string)) >> joinval + + char = someToken('Char') >> tokval + space = someToken('Space') >> tokval + comma = someToken('Comma') >> tokval + quote = someToken('Quote') >> tokval + escape = someToken('Escape') >> tokval + equal = someToken('Equal') >> tokval + true_value = (someCharValue("true") | someCharValue("t") | + someCharValue("True") | someCharValue("TRUE") | someCharValue("T")) + false_value = (someCharValue("false") | someCharValue("f") | + someCharValue("False") | someCharValue("FALSE") | someCharValue("F")) + + escape_space = skip(escape) + space >> joinval + escape_comma = skip(escape) + comma >> joinval + escape_equal = skip(escape) + equal >> joinval + escape_quote = skip(escape) + quote >> joinval + escape_escape = skip(escape) + escape >> joinval + + plain_int_text = someToken('Int') >> tokval + plain_int = plain_int_text >> (lambda v: int(v)) + plain_float_text = someToken('Float') >> tokval + plain_float = plain_float_text >> (lambda v: float(v)) + + identifier = many(char | plain_float_text | plain_int_text | + escape_space | escape_comma | escape_equal | + escape_escape | plain_int_text | quote) >> joinval + quoted_text_ = many(escape_quote | space | plain_int_text | + plain_float_text | char | comma | + escape) >> joinval + quoted_text = skip(quote) + quoted_text_ + skip(quote) + unquoted_text = many(escape_space | escape_comma | + escape_equal | escape_escape | + plain_int_text | char | quote) >> joinval + boolean_value = (true_value >> (lambda s: True) + | false_value >> (lambda s: False)) + + kv_value = plain_int | plain_float | quoted_text | boolean_value + kv = (quoted_text | unquoted_text) + skip(equal) + kv_value >> \ + (lambda x: (x[0], x[1])) + + tag = identifier + skip(equal) + identifier >> (lambda x: (x[0], x[1])) + + def setter(obj, propert): + def r(val): + setattr(obj, propert, val) + return (propert, val) + + return r + + tags = many(skip(comma) + tag) >> (lambda x: x) + fields = (kv + many(skip(comma) + kv)) >> \ + (lambda x: [x[0]] + x[1]) + + write = client.Line(None, None, None, None) + toplevel = (identifier >> setter(write, "key")) + \ + maybe(tags >> setter(write, "tags")) + \ + (skip(space) + (fields >> setter(write, "fields"))) + \ + maybe(skip(space) + plain_int >> setter(write, "timestamp")) + \ + skip(finished) >> (lambda x: x) + + result = toplevel.parse(LineParser.tokenize(line)) + # pprint(result) + return write |