* First shot at improving roman numeral IA.
Added a new converter.
Added new default settings.
Added some tests.
* Added un upper bound on arabic numbers.
* Remove some typo…
* Added some new triggers.
Allowed uncorrect roman number to load the UI anyway.
Fixed a test.
* Added new triggers.
Fixed regular expressions involved in the UI configuration.
Added new tests.
* Add tests to drive development.
Check in work.
Fixed temp triggering ambiguity.
Add experimental conversions US default.
Update the tests.
* Update based on feedback.
* Initial commit.
* Prototype up and running.
* Check in work.
* Update work.
* Checking in work.
* Updates based on feedback.
* Code in design.
* Update padding / margin on controls.
* fix mobile styles
* ColorCodes: Triggers on any query ending in "color <name>"
- Added logic to prevent triggering when excess words are involved in
the sentence that are not expected
Fixes#3294
* - Fixed regex for filler words to be simpler
- Fixed filler count return statement by removing the double negative
* Simplified regex to catch filler words
* Fixed whitespace changes
* Added more tests for when the IA should not trigger. Existing tests already sufficiently cover when the IA should trigger.
* Add sed command example
* Move to file section
* Don't trigger log without brackets.
Quick fix for issue #4246
* Remove test that uses log(10)
* Update Calculator.t
* dont trigger ln without brackets either
* remove ln and log without bracket tests
* previously removed wrong test bring it back
* Using detected number style to make sure queries use a . for a decimal and a , to seperate thousands
* Now using method in number styler to prepare numbers for computation.
* Allow celcisu typo, updated tests accordingly
* Store regex captures in variables
* Allow single unit inputs through guard
* Pass through single unit input to frontend
* Explcitly match guard regex, better check for valid result
* Cleanup whitespace and indentation
* Lowercase symbols that are case-insensitive
* Fix and simplify guard regex
* Better handling of named captures, ensure left unit is not only match
* Case sensitive match first, fall back to lowercase match on symbol
* fix indentation
* remove redundant m on regex match
* Removed default if duplicate.
* Add triggers to list.
* Remove smart match.
* Stash.
* Broaden triggering.
* Update test formatting.
* Revert "Update test formatting."
This reverts commit 390186743c07bffa5d57025e04926d2473a3ff2e.
Reverting because this creates big conflicts in the PR. We can address
the formatting later
* simply language trigger loading
* Add acronyms for teaspoon
* remove trailing spaces
* Add test for abbreviated teaspoon and tablespoon
* remove trailing spaces
* Fix math.js simplify issue.
* Remove debug statements.
* Adds fact to the mix.
* Stash
* Fix greedy quantifier in trig and log/ln
* Don't trigger on anything with stuff after =
* allow triggering on <func> <number>
* Both sides of =
* Fix expressions.
* Tighten up triggering based on valid expressions.
* Remove dumper obj
* Tighten up word triggering.
* Allow only one currency through.
* Kill off symbols in js if allowed through.
* Allow negative numbers.
* Add support for negated numbers.
* Fix initialize issue.
* Refactor code slightly
* Tightened regexs
* Require space between log and input
* Timer: Changed query_raw to query_lc
* also fixes#4188 (regression from ambiguous triggers)
* Updated regex and also solve issue #4188
* Added one more test case
* adding triggers
* indenting key value pairs
* using smart match
* removing liq vol for now
* shortening regex
* adding test
* removing earlier test but stil it fails
* update test
* formatting test
* PrimeFactors: Improve triggering
- Added check to regex guard if query matches form "is <number> prime"
- Strips all preceeding non-digit characters from remainder
Fixes#2799
* Added support for new triggers:
- 'is <number> a prime'
- 'is <number> a prime number'
- 'is <number> prime number'
* ColorPicker: Remove suffixed triggers
* ColorPicker: Handle extra words in remainder
* ColorPicker: Adds tests to check extra words in query
* ColorPicker: Add 'conversion' to remove from query
* ColorPicker: Add test for 'conversion' keyword in remainder
* ColorPicker: Simplify regex, lowercase query
* ColorPicker: Remove print statements
* ColorPicker: Improve triggering
* ColorPicker: Add tests for new triggers
* ColorPicker: Move triggers to separate file
* ColorPicker: Add more combinations of triggers as test
* Geometry Broken Hover and missing calc
* Linked hover effect for dot and SVG
* updated css
* started test file
* consolidated objectInfo YML and finished testing
* removed dumeprs
* fixed triggering and added more triggering test cases
* Implement changed design, fix#3500
* Remove comment
* Update test file to use ISO8601 dates and avoid testing of Role
* Include subtitle in the template
* Take current time using momentjs, calculate difference in JS
* Remove subtitle from the template, make corrections to endCountdown
* Pass date string to JS and find difference in moment objects, change toggling at countdown end
* Remove use of difference passed from the backend
* Update test file
* Begin deprecating perl backend.
* Deprecate backend [skip ci]
* Routine checkin of code.
* Added docs + preprocessing.
* Deprecate perl eval backend.
* Checking in some code.'
* Handle bad search expressions by defaulting to 0
* Add ability to recover with grace from bad expressions from search bar.
* Removed unnessesary perl code.
* Add support for factorials with !
* cleanup trailing spaces
* Removed trig constants.
* Support tau in expression.
* Removed ln preprocess.
* Support leading =. Removed moot code.
* Add support for euro and pound.
* Removed nat lang query parsing from js.
* strip duplicate operators
* Fixed legacy spacing issue in tests.
* Remove empty exponent.
* Bail on hex looking entries.
* better handling of repeated operators
* Remove max import
* revert back to old check for 3+ operators
* relax phone number regex for now
* remove silly test
* added html entities goodie
* fixed trigger
* added tests
* fixed review issues
fixed CSS classes, fixed file names, switched from JSON to YML
* fixed tests and file loading
* renamed to camel case
* deleted AsciiTable.pm
* changed to camel case
* Remove extra newlines
* Fix indentation for Beta server
* Use chompContent
* Simplify CSS
* Remove scrolling div
* Fix test
* Switch overflow control to .c-base__content div
* Add: Initialized ASCII Table Goodie
* Update: Fixed content options
* Update: Title added to td tags
* Update: CSS for tables
* Fix: Makes titles more verbose in template
* Update: Tests for Ascii Table added
* AsciiTable: Fix variable name to be more generic
* AsciiTable: Improved testing of IA
* AsciiTable: Chomp the table
* AsciiTable: Add meta source
* AsciiTable: Rename table_rows -> table
* AsciiTable: Use already existing CSS
* AsciiTable: Data now contains Headers
* AsciiTable: Makes template more generic
* AsciiTable: Updated tests for checking table headers
* AsciiTable: Fix required table styling
* AsciiTable: Fix tests
* Use CSS to set chomp height
* Remove table scroll div
* Switch to List template
* Update AsciiTable.t
* DateRole: Adding failing test for `$nth of $month $year`
* DateRole: Now parses "7th of march 2017"
* DateRole: Adding failing test for '5 of jan 1993' (no suffix and short month)
* DateRole: Handles 5 of march
* DateRole: Now also handle "march the 30th, 2017"
* Conversions: Unphysical units like millibits (mb) and millibytes(mB) should be assumed as mega instead
* Conversions: Now has multiple symbols (case sensitive matches) meaning that Kb and kb both mean kilobits and KB and kB both mean kilobytes :D
* Added variable for non-stribbable triggers.
* Added unstripped triggers to the trigger list
* Added tests to Timer in correspondence of the new triggers
* Removed unnecessary test, added trigger words of hrs and secs and their counterparts.
* Typo fix
* FIX FOR ISSUE #4028
-> Set triggers to only apply to start of search string so the IA isn't triggered for
programming-related searches
-> Removed triggers 'characters count', 'length of string' and 'chars'
as they are unintuitive and can clash with programming requests
-> Added 'of' as a leading word to be removed so queries such as 'length in chars of foo'
count only the characters in 'foo'
-> Updated test script to reflect these changes
* Fix#4028 - reduce overtriggering of 'Chars' goodie
-> Set triggers to only apply to start of search string so the IA isn't triggered for
programming-related searches
-> Removed triggers 'characters count', 'length of string' and 'chars'
as they are unintuitive and can clash with programming requests
-> Added 'of' as a leading word to be removed so queries such as 'length in chars of foo'
count only the characters in 'foo'
-> Updated test script to reflect these changes
* NumbersRole: Despite the role being case insensitive, the number_style_regex isn't! Adding failing test
* NumbersRole: Fixing indentation
* Conversions: Adding tests around implicit case insensitivity in current implementation
* Conversions: Moved all case insensitive behaviour into the handle query so it's explicit rather than relying on query_lc to do the work for us
* Conversions: Adding symbol definitions for case sensitive handling
* Conversions: Fixing markup
* Conversions: UTF8 with BOM on ratios and adding symbol as a trigger
* Conversions: Symbol is now used (case sensitive) to match for factors
* Conversions: The test 'BTU to KwH' should fail (this isn't a unit), however it was causing a division by zero as get_matches wasn't finding a match and this condition wasn't handled correctly
* Conversions: Typo! KwH shouldn't trigger
* Conversions: Tests involving symbols now need to be correct according to SI
* Conversions: Tests around "oz" that should be interpreted as fl oz if the answer should unambiguously be a volume
* Conversions: Will handle ambiguity around oz for more units
* Conversions: Uncommenting tests around digital storage units and fixing them so they're actually correct
* Add Timer files
* Rename Spice -> Goodie
* Move structured answer into Perl
* Update display
Now working.
I removed the 'isMobile' flag, as 'DDG.device.isMobile' was causing it
to crash, and the mobile display wasn't pretty anyway ("Add Timer" was
wrapping). Instead it will use '+' for both.
* Move time parsing into Perl
This way we can check that parsed values are correct in the tests
easily; it also gives us something sensible to return (the start time in
seconds).
* Fix issue with queries like '5s'
* Fix alarm not being located correctly
It was pointing at 'share/spice'.
* Re-enable mobile text
Fixed the issue with mobile display.
* Integrate changes from Spice #3134
Adds the changes from duckduckgo/zeroclickinfo-spice#3134.
* Integrate changes from Spice #2656
Adds the changes from duckduckgo/zeroclickinfo-spice#2656
* Update ID to 'timer_goodie'
* Ensure time from query is used as default
* NumbersRole: No tests covering exponentials
* NumbersRole: Improving test coverage
* NumbersRole: Despite the role being case insensitive, the number_style_regex isn't! Adding failing test
* NumbersRole: number_style_regex now case insensitive
* NumbersRole: Fixing indentation
* Conversions: User feedback has indicated a preference for 3 sig figs for small(ish) results: #3984
* Conversions: Introducing the ability to have significant figures rather than decimal places for numbers <1
* Conversions: The reality is users want to see decimal places for 'normal numbers' (1,765.432 for example) but sig figs for something around 1 so that fractions can be expressed (0.0625 for example)
* Conversions: Bringing tests inline with 3 SF for results < 1; this is more aesthetically pleasing in my opinion
* ParcelForce: Should not trigger for non-UK parcels
* ParcelForce: Starting with KB clashes with Microsoft KB Articles; AFAIK this isn't a valid parcelforce tracking number anyway
* ParcelForce: Working around #3986
* Conversions: The precision (i.e. number of digits after a .) and the significant digits displayed in scientific notation were the same value; there are a few issues with people requesting more accuracy in scientific notation, and none for accuracy
* Conversions: Increasing accuracy of values in tests in-line with new significant figures output
* Conversions: Adding lightyear and lightyears as accepted variants of "light year"
gargaml
PJ Hampton
Philip Kirkbride
Lisa Georgiades
Kyle Daruwalla
Joshua Liu
Billy Barbaro
Aastha
Zaahir Moolla
Rob Emery
Robert Badea
Shivam Shukla
Aditya Tandon
Owen Hines
Manraj Singh
PJ Hampton
Brian Singer
Zaahir Moolla
Rasika Pohankar
lernae
Emily Tackett
Nalin Bhardwaj
TheSpeedyBlur
Sophie Durrant
Ed Callaghan
Ben Moon
Shady Tantawy
adv16
Jag Talon