ó
šßÈ[c @` sl d Z d d l m Z m Z m Z d d l Z d d l Z d d l m Z d d l m
Z
m Z d d l m
Z
d d l m Z d d
l m Z d d l m Z d e f d
„ ƒ YZ d d d „ ƒ YZ d „ Z d e
j f d „ ƒ YZ d e
j f d „ ƒ YZ d e
j f d „ ƒ YZ d e
j f d „ ƒ YZ d e
j f d „ ƒ YZ! d e
j" f d „ ƒ YZ# d S( sÇ An extensible HTML table reader and writer.
html.py:
Classes to read and write HTML tables
`BeautifulSoup `_
must be installed to read HTML tables.
i ( t absolute_importt divisiont print_functionNi ( t six( t zipt rangei ( t core( t Column( t writer( t deepcopyt
SoupStringc B` s e Z d Z d „ Z d „ Z RS( s8
Allows for strings to hold BeautifulSoup data.
c O` s t j | | | Ž S( N( t strt __new__( t clst argst kwargs( ( s4 lib/python2.7/site-packages/astropy/io/ascii/html.pyR s c C` s
| | _ d S( N( t soup( t selft val( ( s4 lib/python2.7/site-packages/astropy/io/ascii/html.pyt __init__" s ( t __name__t
__module__t __doc__R R ( ( ( s4 lib/python2.7/site-packages/astropy/io/ascii/html.pyR
s t
ListWriterc B` s e Z d Z d „ Z d „ Z RS( sD
Allows for XMLWriter to write to a list instead of a file.
c C` s
| | _ d S( N( t out( R R ( ( s4 lib/python2.7/site-packages/astropy/io/ascii/html.pyR + s c C` s | j j | ƒ d S( N( R t append( R t data( ( s4 lib/python2.7/site-packages/astropy/io/ascii/html.pyt write. s ( R R R R R ( ( ( s4 lib/python2.7/site-packages/astropy/io/ascii/html.pyR & s c C` s‹ | d k s | j d k r t Sd | k r5 | d k S| d } t | t j ƒ rn d | j k om | d | k St | t ƒ r‡ | | k St S( sb
Checks whether the given BeautifulSoup tag is the table
the user intends to process.
t tablet table_idi t idN( t Nonet namet Falset
isinstanceR t string_typest attrst int( R t htmldictt numtableR ( ( s4 lib/python2.7/site-packages/astropy/io/ascii/html.pyt identify_table2 s
t HTMLInputterc B` s e Z d Z d „ Z RS( s›
Input lines of HTML in a valid form.
This requires `BeautifulSoup
`_ to be installed.
c C` sx y d d l m } Wn t k
r6 t j d ƒ ‚ n Xd | j k r t j ƒ * t j d d ƒ | d j | ƒ ƒ } Wd QXn | d j | ƒ | j d ƒ } | j
d ƒ } xš t | ƒ D]/ \ } } t | | j | d
ƒ r¼ | } Pq¼ q¼ Wt
| j d t ƒ r d j | j d ƒ } n d
j | j d ƒ } t j d j | ƒ ƒ ‚ g | j
d ƒ D] } t | ƒ ^ q\}
|
S( sh
Convert the given input into a list of SoupString rows
for further processing.
i ( t
BeautifulSoups3 BeautifulSoup must be installed to read HTML tablest parsert ignores& .*no parser was explicitly specified.*s
NR i R s
number {0}s id '{0}'s ERROR: HTML table {0} not foundt tr( t bs4R* t ImportErrorR t OptionalTableImportErrort htmlt warningst catch_warningst filterwarningst joint find_allt enumerateR( R" R% t formatt InconsistentTableErrorR
( R t linesR* R t tablest it possible_tableR t err_descrt xt soup_list( ( s4 lib/python2.7/site-packages/astropy/io/ascii/html.pyt
process_linesP s*
(( R R R RA ( ( ( s4 lib/python2.7/site-packages/astropy/io/ascii/html.pyR) H s t HTMLSplitterc B` s e Z d Z d „ Z RS( s
Split HTML table data.
c c` sú xÏ | D]Ç } t | t ƒ s+ t d ƒ ‚ n | j } | j d ƒ } | r• g | D]= } | j d ƒ r~ | j j ƒ | d f n | j j ƒ ^ qP Vn | j d ƒ } | r g | D] } | j j ƒ ^ q± Vq q Wt | ƒ d k rö t j
d ƒ ‚ n d S( s=
Return HTML data from lines as a generator.
s' HTML lines should be of type SoupStringt tht colspant tdi s. HTML tables must contain data in a tagN( R" R
t TypeErrorR R6 t has_attrt textt stript lenR R9 ( R R: t lineR t header_elementst elt
data_elements( ( s4 lib/python2.7/site-packages/astropy/io/ascii/html.pyt __call__{ s
I(( R R R RO ( ( ( s4 lib/python2.7/site-packages/astropy/io/ascii/html.pyRB v s t
HTMLOutputterc B` sY e Z d Z e j e j ƒ e j e j ƒ e j e j ƒ e j e j ƒ g Z
d „ Z RS( sÍ
Output the HTML data as an ``astropy.table.Table`` object.
This subclass allows for the final table to contain
multidimensional columns (defined using the colspan attribute
of | ).
c C` sÜ g } d } x´ | t | ƒ k r | | } t | d ƒ r¨ | | | | j !} t j | j ƒ } t t g | D] } | j ^ qm Œ ƒ | _ | j | ƒ | | j 7} q | j | ƒ | d 7} q Wt
t | ƒ j | | ƒ S( s?
Process the data in multidimensional columns.
i RD i (
RJ t hasattrRD R R R t listR t str_valsR t superRP RO ( R t colst metat new_colst col_numt colt span_colst new_colR? ( ( s4 lib/python2.7/site-packages/astropy/io/ascii/html.pyRO ž s
+
( R R R R t
convert_numpyt numpyR% t floatR t unicodet default_convertersRO ( ( ( s4 lib/python2.7/site-packages/astropy/io/ascii/html.pyRP s t
HTMLHeaderc B` s e Z e Z d „ Z d „ Z RS( c C` s[ xT t | ƒ D]F \ } } t | t ƒ s7 t d ƒ ‚ n | j } | j d k r
| Sq
Wd S( sE
Return the line number at which header data begins.
s' HTML lines should be of type SoupStringN( R7 R" R
RF R RC R ( R R: R< RK R ( ( s4 lib/python2.7/site-packages/astropy/io/ascii/html.pyt
start_line¸ s c C` sÿ g | _ g } xà | j D]Õ } t | t ƒ rÅ t j d | d ƒ } t | d ƒ | _ | j j | ƒ | j | d ƒ xs t d t | d ƒ ƒ D], } | j j t j d ƒ ƒ | j d ƒ q’ Wq | j j t j d | ƒ ƒ | j | ƒ q W| | _ d S( sU
Set columns from header names, handling multicolumns appropriately.
R i i t N(
RU t namesR" t tupleR R R% RD R R ( R t new_namesR RY R< ( ( s4 lib/python2.7/site-packages/astropy/io/ascii/html.pyt _set_cols_from_namesÆ s ( R R RB t splitter_classRb Rg ( ( ( s4 lib/python2.7/site-packages/astropy/io/ascii/html.pyRa µ s t HTMLDatac B` s e Z e Z d „ Z d „ Z RS( c C` s‹ xu t | ƒ D]g \ } } t | t ƒ s7 t d ƒ ‚ n | j } | j d k r
| j d k rp t j d ƒ ‚ n | Sq
Wt j d ƒ ‚ d S( sD
Return the line number at which table data begins.
s' HTML lines should be of type SoupStrings9 HTML tables cannot have headings and data in the same rows! No start line found for HTML dataN(
R7 R" R
RF R RE R RC R R9 ( R R: R< RK R ( ( s4 lib/python2.7/site-packages/astropy/io/ascii/html.pyRb á s c C` sz d } xY t | ƒ D]K \ } } t | t ƒ s= t d ƒ ‚ n | j } | j d k r | } q q W| d k rr d S| d S( sB
Return the line number at which table data ends.
iÿÿÿÿs' HTML lines should be of type SoupStringi N( R7 R" R
RF R RE R ( R R: t
last_indexR< RK R ( ( s4 lib/python2.7/site-packages/astropy/io/ascii/html.pyt end_lineó s
( R R RB Rh Rb Rk ( ( ( s4 lib/python2.7/site-packages/astropy/io/ascii/html.pyRi Þ s t HTMLc B` sb e Z d Z d Z d g Z d Z d Z e Z e Z
e Z i d „ Z
d „ Z d „ Z d „ Z RS( sy
Read and write HTML tables.
In order to customize input and output, a dict of parameters may
be passed to this class holding specific customizations.
**htmldict** : Dictionary of parameters for HTML input/output.
* css : Customized styling
If present, this parameter will be included in a |