Help:Sorting

From SacredWiki
Jump to navigation Jump to search
MediaWiki Handbook: Contents, Readers, Editors, Moderators, System admins +/-

Tables can be made sortablehttp://en.wikipedia.org/index.php/Sorting via client-side JavaScript with class="sortable" (in combination with the usual formatting: class="wikitable sortable"). This works in MediaWiki 1.9 and above, which is installed in all Wikimedia projects.

A sortable table is identified by the arrows in each of its header cells. Clicking them will cause the table rows to sort based on the selected column, in ascending order first, and subsequently toggling between ascending and descending order. Links and other wiki-markup are not possible in headers.

JavaScript

The JavaScript code jquery.tablesorter.js (source) of the tablesorter is loaded by the ResourceLoader. Some sites may have a page MediaWiki:Common.js which adds and overrides some code. Browsers need to support JavaScript and it needs to be enabled for sorting to work.

Sort modes

As of version 1.16.5, the way items are sorted depends on the data type of the item currently in the first row. This is true for the top cell of the column in both ascending and descending order. To determine the data type, multiple cells are tested and the most appropriate format is chosen. Mismatches are possible. The sort order of a column can be forced. See the relevant section farther down.

Tags such as span or sup are ignored when determining data type.

Dates

Various date formats are supported, including those with localized month names. On the German Wikipedia, "16. März 2010" is correctly sorted as 2010-03-16

Most other numerical formats are supported as well, including those with different separators (such as . , ' or / ); On English Wikipedias dates are treated as US-Dates (eg. month-day-year) per default.

Numbers

The script can recognize numbers with different decimal separators (. and ,) as well as e/E numbers. However, numbers will be sorted alphanumerically (with 9 sorted after 10) unless this default behaviour is overridden. (See below.)

Text

Text is sorted in ASCII order (Any accented/special characters follow after the basic latin alphabet). This can be changed site wide by posting code like the following inside the common.js:

mw.config.set('tableSorterCollation', {'ä':'ae', 'ö' : 'oe', 'ß': 'ss', 'ü':'ue'});

Afterwards, all 'ä' will be sorted as if they were an ae etc. Partial list showing the default order: !"#$%&'()*+,-./09:;<=>?@[\]^_'az{|}~é—

Forcing the sort mode for a column

The sort mode can be manually specified by putting data-sort-type inside the header of the respective row. This functionality is based on tablesorter.com. The following (case-insensitive) values are valid for data-sort-type:

  • text
  • number
  • IPAddress
  • currency
  • url
  • isoDate
  • usLongDate
  • date
  • time

Example:

{|class="wikitable sortable"
! data-sort-type="date" | Date!!Name!!Height
|-
|01.10.1977||Smith||1.85
|-
|11.6.1972||Ray||1.89
|-
|1.9.1992||Bianchi||1.72
|}
Date Name Height
01.10.1977 Smith 1.85
11.6.1972 Ray 1.89
1.9.1992 Bianchi 1.72

Examples

The first example demonstrates that text is positioned at zero, and that e.g. e3 for 1000 is not allowed; use 1e3 instead. It also shows that "-" should be used, not "−" (a minus sign).

The second example shows that expressions are not sorted according to their evaluated value, but according to the first number.

The third example shows that a percentage is accepted for numeric sorting mode, but ignored in the actual sorting, so if a column contains percentages, all numbers have to be written as a percentage.

The fourth example shows again that "ca. 12" sorts at 0, as opposed to 12 with some text after it, which sorts at 12. In case such an element arrives at the top of a column, it causes alphabetic sorting mode.

numbers
 
-
4.0
15
10
1 2
1 aa
1 a b
1 aa
1 11
1 1 a
1 1
1
192
123,456.789
123,456,789
2,500,000,000
300,000,000
3,000,000 abc
5,000,000
2,000 def
-4,000
aaa
-9,999
4,000
9,999
800,000
900,000
numbers
123 564,589.7e12
9
-80
80 abc 5
abc 80
70
600
first alphabetic, later also numeric mode
123.4 ghi
2,500,000,000
300,000,000
3,000,000 abc
5,000,000
2,000 def
4,000
9,999
800,000
900,000
currencies
$ 9
$ 80
$ 70
$ 600
currencies
€ 9
€ 80
€ 70
€ 600
currencies
£ 9
£ 80
£ 70
£ 600
currencies
¥ 9
¥ 80
¥ 70
¥ 600
comparison
a 9
a 80
a 70
a 600
comparison
e 9
e 80
e 70
e 600

The example with "a" gives alphabetic sorting; that with "e" ditto, the data are not mistaken for numbers in scientific format.

mixed notations
1.4285714285714E+17
1000000000000000000
-1000000000000000000
.0000000000000000001
-.0000000000000000001
-1.4285714285714E+17
1.4285714285714E-13
-1.4285714285714E-13
89 123 456 788
89,123,456,789
333
1e10
e 9
e 80
e 70
e 600
999e9
88e80
7e270
999e-9
88e-80
7e-270
-999e9
−999e9
-88e80
-7e270
-999e-9
-88e-80
-7e-270
e3
-e3
1e3
e9
e80
e270
6e11
8e11
first number in each element counts
7-4
2
4
22/7
111
percentage
7%
2
4
22
111
mixed notations
14
-14
11
-12 (retrograde)
12 or 13
12 (?)
ca. 12
12 (approx.)
?

Additional features

Excluding the last row from sorting

Sometimes it is helpful to exclude the last row of a table from the sorting process.

This can be achieved by declaring the last row as a footer

Wiki markup

{|class="wikitable sortable"
!Name!!Surname!!Height
|-
|John||Smith||1.85
|-
|Ron||Ray||1.89
|-
|Mario||Bianchi||1.72
|-
! !!Average:||1.82
|}

What it looks like in your browser

Name Surname Height
John Smith 1.85
Ron Ray 1.89
Mario Bianchi 1.72
Average: 1.82

Excluding the first row from sorting

The same can be applied for first rows as well, by declaring them as header using the same exclamation mark notation.

Name Surname Height
Average: 1.82
John Smith 1.85
Ron Ray 1.89
Mario Bianchi 1.72
Average: 1.82

Making a column unsortable

If you want a specific column not to be sortable, specify class="unsortable" in the attributes of its header cell.

Wiki markup

{|class="wikitable sortable"
!Numbers!!Alphabet!!Dates!!Currency!!class="unsortable"|Unsortable
|-
|1||Z||02-02-2004||5.00||This
|-
|2||y||13-apr-2005||||Column
|-
|3||X||17.aug.2006||6.50||Is
|-
|4||w||01.Jan.2005||4.20||Unsortable
|-
|5||V||05/12/2006||7.15||See?
|-
!Total: 15!!!!!!Total: 29.55!!
|-
|}

What it looks like in your browser

Numbers Alphabet Dates Currency Unsortable
1 Z 02-02-2004 5.00 This
2 y 13-apr-2005 Column
3 X 17.aug.2006 6.50 Is
4 w 01.Jan.2005 4.20 Unsortable
5 V 05/12/2006 7.15 See?
Total: 15 Total: 29.55 Original example

Specifying a sort key

Sometimes the value of a cell is not correctly parsed or one wants to sort the row in a special way. (e.g. a cell containing 'John Doe' should actually be sorted as 'Doe' and not as 'John') This can be easily achieved by setting the data-sort-value attribute.

Note, however, that this makes use of a new feature in HTML5, which is enabled by default in MediaWiki (including WMF wikis since September 2012 cfr. bugzilla:27478).

Wiki markup

{|class="wikitable sortable"
!Name and Surname!!Height
|-
|data-sort-value="Smith, John"|John Smith||1.85
|-
|data-sort-value="Ray, Ian"|Ian Ray||1.89
|-
|data-sort-value="Bianchi, Zachary"|Zachary Bianchi||1.72
|-
!Average:||1.82
|}

This gives:

Name and Surname Height
John Smith 1.85
Ian Ray 1.89
Zachary Bianchi 1.72
Average: 1.82

See also mediawiki.org.

Keeping some rows together

data-sort-value can be used to keep certain rows together. The original mutual order of these rows is preserved.

Example where this is the case for the rows about the Netherlands:

Country/province Capital
France Paris
Netherlands Amsterdam
South Holland The Hague
U.K. London

Special dates

For years BC we can use, for example, !9937-09-23 for -0062-09-23 (subtract the year number BC from 10000, or the absolute value of the astronomical year from 9999).

If a table column contains any or all incomplete dates, this will not cause sorting problems. If only a year and month are given, that incomplete date is positioned alphabetically before the first day of the month in question. Likewise, if only a year is given, the date is positioned before the first month or day given for that year.

Use of #time

Using parser function #time we can put <span style="display:none">&{{#expr:3e11+{{#time:U|..}}}}</span> in front of the displayed date. This works in the range 1 Jan 111, 00:00:00 through 31 Dec 9999, 23:59:59 for the proleptic Gregorian calendar. The added value makes all values positive and the same length (if scientific format would show up an additional step is needed to prevent this). The "&" forces string sort mode.

Dates and times can be entered in any php date/time format. Note that when we have just a year, a month (typically Jan) must be added in the hidden part.

Example using Help:Sorting/date:

input date text date and time as interpreted, with hidden sortkey input with visible sortkey input with hidden sortkey Unix time

Help:Sorting/date

Help:Sorting/date Help:Sorting/date Help:Sorting/date Help:Sorting/date Help:Sorting/date Help:Sorting/date Help:Sorting/date Help:Sorting/date Help:Sorting/date Help:Sorting/date Help:Sorting/date Help:Sorting/date Help:Sorting/date Help:Sorting/date Help:Sorting/date Help:Sorting/date Help:Sorting/date Help:Sorting/date Help:Sorting/date Help:Sorting/date Help:Sorting/date Help:Sorting/date Help:Sorting/date Help:Sorting/date Help:Sorting/date Help:Sorting/date Help:Sorting/date Help:Sorting/date Help:Sorting/date Help:Sorting/date Help:Sorting/date Help:Sorting/date Help:Sorting/date Help:Sorting/date Help:Sorting/date Help:Sorting/date Help:Sorting/date Help:Sorting/date Help:Sorting/date Help:Sorting/date Help:Sorting/date Help:Sorting/date Help:Sorting/date Help:Sorting/date Help:Sorting/date Help:Sorting/date Help:Sorting/date Help:Sorting/date Help:Sorting/date Help:Sorting/date Help:Sorting/date

To use dates before the year 111, add a multiple of 400, e.g. 6000, to all years, this effectively shifts the range to 1 Jan -5889, 00:00:00 through 31 Dec 3999, 23:59:59, without changing the calendar.

See also:

Secondary sortkey

It is possible to sort by column A (primary sortkey), while for equal values in column A, sort by column B (secondary sortkey): first sort by A by clicking the sort button of column A once or twice, then, while holding the shift-key, click the sort button of column B once or twice.

Example:

First click on column Text and then, while holding the shift-key, on Numbers, you'll see that the ordering is on Text (1), Numbers (2).

Numbers Text Dates Currency More text
4 a 01.Jan.2005 4.20 row 1
5 a 05/12/2006 7.15 row 2
1 b 02-02-2004 5.00 row 3
1 a 02-02-2004 5.00 row 4
2 x 13-apr-2005 row 5
2 a 13-apr-2005 row 6
3 a 17.aug.2006 6.50 row 7
3 z 25.aug.2006 2.30 row 8
3 z 28.aug.2006 5.50 row 9
3 z 31.aug.2006 3.77 row 10
3 z 01.sep.2006 1.50 row 11
Bottom

Cell spanning multiple rows/cells

Cells which are spanning more than one row are treated as if it were multiple cells with the same value.

Example:

Date Name Height
01.10.1977 Smith 1.85
11.6.1972 Adams
1.9.1992 Bianchi 1.72


Javascript sorting may not work properly on tables with cells extending over multiple rows and/or columns (however, sorting of columns up to and including the first with colspan does not seem to be affected). Also, while cells can be empty, they should not be missing at the end of a row. In these cases sometimes the table gets messed up when attempting to sort, while other times some of the sorting buttons work while others don't.

Colspan workaround

To allow sorting, the formal number of cells in each row should be equal (if not all columns are made sortable this should apply at least for the number of cells up to and including the last sortable column). However, with a CSS hack the number of cells shown in a row can differ from the formal number of cells. For example, two formal cells can be shown as one by specifying a width for the first column, shifting the contents of the second cell to the left, increasing its width by the same amount, and hiding the cell border that would normally be visible. Hidden sortkeys can be used to control, for sorting with respect to each column, how this row should be sorted.

Example:

Country Capital
France Paris
Z M
Sorting with respect to the first column this row sorts like Z, with respect to the second column like M
U.K. London

This can be combined with the method of "keeping some rows together" demonstrated above. For an example of an application of this, consider a table of three columns where the third column would make the table too wide, such as a column of miscellaneous details. These details can be put in separate rows, each staying below the corresponding row when the table is sorted.

Example:

Country Capital
France Paris
In Paris is the Eiffel Tower.
France Paris
In Paris is the Eiffel Tower.
U.K. London
In the U.K. you cannot pay with euros.
U.K. London
In the U.K. you cannot pay with euros.
Germany Berlin
Germany includes the former DDR.
Germany Berlin
Germany includes the former DDR.

A table row template makes this technique less cumbersome to apply, see e.g. w:List of furry conventions, w:Template:Furry-con-list-start and w:Template:Furry-con-list-entry.

Controlling sorting and display

Text undesired for sorting but needed for display:

  • In numeric sorting mode, the sorting will still work properly even though the cell (except the cell of the first data row) contains text after numbers (e.g. "200 approx"). Empty cell is treated as "zero" when sorting numerically. See e.g. Help:Sorting/countries. However bear in mind that the cell of the first data row will change accordingly after sorting. If that cell no longer contains number only after sorting, the sorting mode will change. For example, if the cell of the first data row becomes "200 approx" after sorting, this will make the sorting mode alphabetic.
  • In date sorting mode, this text needs to be put in a separate column; in the case of a cell containing a range of dates or numbers (e.g. from .. to ..), text in surplus of what is required for sorting is put in the extra column. If the first part of the text is used for sorting, then the extra column needs to be the following one; conversely, if the last part of the text is used for sorting, then the extra column needs to be the previous one; depending on the table format, this dividing of an item over two cells may look ugly.
  • In alphabetic sorting, any footnotes etc. do not require a separate column; they can simply be put at the end of the element.


Static column

A static column, e.g. with row numbers, can be obtained with two side-by-side tables with for each row the same height set in both tables:

Number
1
2
Country Capital
The Netherlands Amsterdam (although The Hague is the seat of government)
France Paris

The style can be adjusted to make it appear as a single table. If for some row the height of that row is too small for the text in a cell on one of the sides, the browser increases it, and there is no longer a match.

Default order

It is not possible to make a table appear sorted by a certain column without the user clicking on it. By default, the rows of a table always appear in the same order as in the wikitext. If you want a table to appear sorted by a certain column, you must sort the wikitext itself in that order; see the next section for one way to do this.

Sorting the wikitext of a table

Sorting the wikitext itself, thus creating a new default sort order, can be done semi-automatically as follows. Take the wikitext of the table without top and bottom lines. Use "find and replace" to replace the cell separators with special code not containing "|". If there are pipes in the table cells, replace all pipes by some code, and replace that code with a newline in front of it (originating from the code for the start of a new row) back. Apply mw:Module:Sort (see mw:Module talk:Sort) at mw:Special:ExpandTemplates by putting: {{#invoke:Sort|f|{{!}}-
{{!}} (with the newline) before, and }} after the wikitext, to sort the items between the pipes, with the desired separator in the result. Discard the items at the start containing "-" and a newline. Restore the cell separators and the pipes in the cells by replacing the temporary codes for them. Readd the top and bottom lines.

This method sorts by the wikitext of the rows, so in principal by the first column (and the second as secondary key), although wikitext codes in the cells of the first column before the content can affect the order.

Basic alphabetic sorting order

demo
!
"
#
$
%
&
'
(
)
*
+
,
-
.
/
0
9
:
;
<
=
>
?
@
[
\
]
^
_
`
A
Z
a
z
A1
Z1
a1
z1
{
|
}
~
É
é
É1
é1

The two-character entries such as A1 demonstrate that A and a are at the same position.

This is not a fully alphabetic sort order: letter case is first folded to lowercase using a basic 1-to-1 conversion table (limited to the Basic Multilingual Plane of Unicode, and whose coverage and completeness still depends on browser versions and on their current implementation of the versioned Unicode Character Database), but letters with diacritics (and all other digits, symbols or special whitespaces or format control characters) will still sort according to the binary encoding of the casefolded letter, using the binary order of the UTF-16 code units (exposed and seen in Javascript through the parsed HTML DOM), but not the binary order of UTF-8 code units in the HTML page, and not of codepoints as one could also expect for encoded characters in supplementary planes).

In addition, no normalization of the Unicode text is being performed (so canonically equivalent strings, that should compare equal or with only very minor binary difference, may sometimes compare very far away, with completely different strings interleaved between them). For this reason, MediaWiki pages should always be encoded with their text in the Normalized Form C (preComposed), as recommended in the HTML standards.

As of today, an UCA-based sort is still not implemented in the client-side Javascript code, but some wikis are implementing a limited form of multilevel collation using custom basic replacement rules tuned for specific languages.

Server issue

It has been observed that the MediaWiki code on the server replaces a regular space before "!" by a non-breaking space &#160;, affecting the sorting order. To avoid this, this blank space can be coded as &#32;, or the exclamation mark may be surrounded by <nowiki> and </nowiki> tags. This is to comply with French typographic rules, where exclamation marks (and a few other punctuations) must be preceded (or sometimes followed) by a space (preferably narrow) which must still be unbreakable when it is effectively needed and present, the substitution being performed as an convenient editing facility of the Wiki code for cases that are very frequent within many texts.

Persistent sort states using cookies

Adding this snippet to your MediaWiki:Common.js page will make the sortable tables remember their columns sort states in a cookie so they look the same next time the page is visited. Each sortable table must have a unique id attribute for its state to be stored in the cookie. <source lang=javascript> addOnloadHook( function() {

   jQuery('.sortable').each( function() {
       var id = jQuery(this).attr('id');
       document.shCookie = getCookie('sortheader-'+id);
       document.sortheaderId = 0;
       jQuery('#'+id+' a.sortheader').each( function() {
           var id = jQuery(this).parent().parent().parent().parent().attr('id');
           var sh = document.sortheaderId++;
           if( sh+100 == document.shCookie ) { ts_resortTable(this); ts_resortTable(this); }
           if( sh == document.shCookie ) { ts_resortTable(this); sh += 100; }
           jQuery(this).bind('click', {id: id, sh: sh}, function(e) {
               setCookie('sortheader-'+e.data.id, e.data.sh, 1);
               e.data.sh += e.data.sh < 100 ? 100 : -100;
           });
       });
   });

});

function setCookie(c_name,value,exdays) {

   var exdate=new Date();
   exdate.setDate(exdate.getDate() + exdays);
   var c_value=escape(value) + ((exdays==null) ? "" : "; expires="+exdate.toUTCString());
   document.cookie=c_name + "=" + c_value;

}

function getCookie(c_name) {

   var i,x,y,ARRcookies=document.cookie.split(";");
   for (i=0;i<ARRcookies.length;i++) {
       x=ARRcookies[i].substr(0,ARRcookies[i].indexOf("="));
       y=ARRcookies[i].substr(ARRcookies[i].indexOf("=")+1);
       x=x.replace(/^\s+|\s+$/g,"");
       if (x==c_name) return unescape(y);
   }

} </source>

See also

Examples elsewhere:


Links to other help pages

Help contents
Meta · Wikinews · Wikipedia · Wikiquote · Wiktionary · Commons: · mw: · b: · s: · mw:Manual · Google
Versions of this help page (for other languages see further)
Meta · Wikinews · Wikipediahttp://en.wikipedia.org/index.php/Help:Sorting · Wikiquote · Wiktionary
What links here on Meta or from Meta · Wikipedia · MediaWiki
Reading
Go · Search · Stop words · Namespace · Page name · Section · Backlinks · Redirect · Category · Image page · Special pages · Printable version
Tracking changes
Recent changes (enhanced) | Related changes · Watching pages · Diff · Page history · Edit summary · User contributions · Minor edit · Patrolled edit
Logging in and preferences
Logging in · Preferences · User style
Editing
Starting a new page · Advanced editing · Editing FAQ · Edit toolbar · Export · Import · Shortcuts · Edit conflict · Page size
Referencing
Links · URLs · Piped links · Interwiki linking · Footnotes
Style and formatting
Wikitext examples · CSS · Reference card · HTML in wikitext · Formula · List · Table · Sorting · Colors · Images and file uploads
Fixing mistakes
Show preview · Testing · Reverting edits
Advanced functioning
Expansion · Template · Advanced templates · Parser function · Parameter default · Variable · System message · Substitution · Array · Calculation · Embed page
Others
Special characters · Renaming (moving) a page · Talk page · Signatures · Sandbox · Legal issues for editors