swh:1:snp:7ce5f1105410d5ee1ad6abfdc873986c25b579e5
Tip revision: d5251b0e56c33692d1d1dcd24fb902e8583e2c14 authored by Dirk Roorda on 28 June 2021, 05:50:26 UTC
python version requirement back to 3.6.3
python version requirement back to 3.6.3
Tip revision: d5251b0
display.py
"""
# Display
Where the advanced API really shines is in displaying nodes.
There are basically two ways of displaying a node:
* *plain*: just the associated text of a node, or if that would be too much,
an identifying label of that node (e.g. for books, chapters and lexemes).
* *pretty*: a display of the internal structure of the textual object a node
stands for. That structure is adorned with relevant feature values.
These display methods are available for nodes, tuples of nodes, and iterables
of tuples of nodes (think: query results).
The names of these methods are
* `plain`, `plainTuple`, and `table`;
* `pretty`, `prettyTuple` and `show`.
In plain and pretty displays, certain parts can be *highlighted*, which is
good for displaying query results where the parts that correspond directly to the
search template are highlighted.
## Display parameters
There is a bunch of parameters that govern how the display functions arrive at their
results. You can pass them as optional arguments to these functions,
or you can set up them in advance, and reset them to their original state
when you are done.
All calls to the display functions look for the values for these parameters in the
following order:
* optional parameters passed directly to the function,
* values as set up by previous calls to `displaySetup()`,
* corpus dependent default values configured by the advanced API.
See `tf.advanced.options` for a list of display parameters.
## Rendering
Both `pretty` and `plain` are implemented as a call to the
`tf.advanced.render.render` function.
## See also
All about the nature and implementation of the display algorithm is in
`tf.about.displaydesign`.
"""
import os
import types
from ..parameters import DOWNLOADS, SERVER_DISPLAY, SERVER_DISPLAY_BASE
from ..core.helpers import mdEsc
from .helpers import getRowsX, tupleEnum, RESULT, dh, showDict
from .condense import condense, condenseSet
from .highlight import getTupleHighlights
from .options import Options
from .render import render
from .unravel import unravel, _getLtr
LIMIT_SHOW = 100
LIMIT_TABLE = 2000
def displayApi(app, silent):
"""Produce the display API.
The display API provides methods to generate styled representations
of pieces of corpus texts in their relevant structures.
The main end-user functions are `plain(node)` and `pretty(node)`.
`plain()` focuses on the plain text, `pretty()` focuses on structure
and feature display.
Related are `plainTuple()` and `prettyTuple()` that work for tuples
instead of nodes.
And further there are `show()` and `table()`, that work
with iterables of tuples of nodes (e.g. query results).
Parameters
----------
app: obj
The high-level API object
silent:
The verbosity mode to perform this operation in.
Normally it is the same as for the app, but when we do an `A.reuse()`
we force `silent=True`.
"""
app.export = types.MethodType(export, app)
app.table = types.MethodType(table, app)
app.plainTuple = types.MethodType(plainTuple, app)
app.plain = types.MethodType(plain, app)
app.show = types.MethodType(show, app)
app.prettyTuple = types.MethodType(prettyTuple, app)
app.pretty = types.MethodType(pretty, app)
app.unravel = types.MethodType(unravel, app)
app.loadCss = types.MethodType(loadCss, app)
app.getCss = types.MethodType(getCss, app)
app.displayShow = types.MethodType(displayShow, app)
app.displaySetup = types.MethodType(displaySetup, app)
app.displayReset = types.MethodType(displayReset, app)
app.display = Options(app)
if not app._browse:
app.loadCss()
def displayShow(app, *options):
"""Show display parameters.
Shows current values of all or selected display parameters.
Parameters
----------
options: keys
Options of which the current value will be shown.
If no option is passes, all options will be shown.
See Also
--------
tf.advanced.settings: options allowed in `config.yaml`
"""
display = app.display
display.setup()
data = display.current
showDict("<b>current display options</b>", data, *options)
def displaySetup(app, **options):
"""Set up all display parameters.
Shows current values of display parameters and/or
assigns working values to display parameters.
All subsequent calls to display functions such as `plain` and `pretty`
will use these values, unless they themselves are passed overriding
values as arguments.
These working values remain in effect until a new call to `displaySetup()`
assigns new values, or a call to `displayReset()` resets the values to the
defaults.
!!! hint "show current values"
The defaults themselves come from the corpus settings, which are influenced
by its `config.yaml` file, if it exists. See `tf.advanced.settings`.
You can show the current values by means of `displayShow`.
Parameters
----------
show: keys
Options of which the current value will be shown.
options: key-values
Explicit values for selected options that act as overrides of the defaults.
See Also
--------
tf.advanced.settings: options allowed in `config.yaml`
tf.advanced.options: all available display options
"""
display = app.display
display.setup(**options)
def displayReset(app, *options):
"""Restore display parameters to their defaults.
Reset the given display parameters to their default value and let the others
retain their current value.
So you can reset the display parameters selectively.
Parameters
----------
options: list, optional `[]`
If present, only restore these options to their defaults.
Otherwise, restore all display settings.
"""
display = app.display
display.reset(*options)
def loadCss(app):
"""Load the CSS for this app.
If we are in the TF-browser, the generic CSS is already provided, we only
need to respond with the app-specific CSS: we return it as string.
The flag `app._browse` is used to steer us into this case.
Otherwise, we collect the complete CSS code from Text-Fabric and the app,
and we add a piece to override some of the notebook CSS for tables,
which specify a table layout with right aligned cell contents by default.
We then load the resulting CSS into the notebook.
Returns
-------
None | string
When in the TF browser, the app-dependent CSS is returned.
Otherwise, nothing is returned, but the complete CSS is displayed as HTML in the notebook.
"""
_browse = app._browse
aContext = app.context
appCss = aContext.css
if _browse:
return appCss
css = getCss(app)
dh(css)
def getCss(app):
"""Export the CSS for this app.
We collect the complete CSS code from Text-Fabric and the app,
and we add a piece to override some of the notebook CSS for tables,
which specify a table layout with right aligned cell contents by default.
Returns
-------
None | string
CSS code, including a surrounding `<style>` element.
"""
aContext = app.context
appCss = aContext.css
cssPath = (
f"{os.path.dirname(os.path.dirname(os.path.abspath(__file__)))}"
f"{SERVER_DISPLAY_BASE}"
)
genericCss = ""
for cssFile in SERVER_DISPLAY:
with open(f"{cssPath}/{cssFile}", encoding="utf8") as fh:
genericCss += fh.read()
tableCss = (
"tr.tf.ltr, td.tf.ltr, th.tf.ltr { text-align: left ! important;}\n"
"tr.tf.rtl, td.tf.rtl, th.tf.rtl { text-align: right ! important;}\n"
)
return f"<style>{tableCss}{genericCss}{appCss}</style>"
def export(app, tuples, toDir=None, toFile="results.tsv", **options):
"""Exports an iterable of tuples of nodes to an Excel friendly `.tsv` file.
!!! hint "Examples"
See for detailed examples the
[exportExcel (bhsa)](https://nbviewer.jupyter.org/github/annotation/tutorials/blob/master/bhsa/exportExcel.ipynb)
and
[exportExcel (oldbabylonian)](https://nbviewer.jupyter.org/github/annotation/tutorials/blob/master/oldbabylonian/exportExcel.ipynb)
notebooks.
Parameters
----------
tuples: iterable of tuples of integer
The integers are the nodes, together they form a table.
The table maybe uniform or not uniform,
which matters to the output. See below.
toDir: string, optional `None`
The destination directory for the exported file.
By default it is your Downloads folder.
If the directory does not exist, it will be created.
toFile: boolean, optional `results.tsv`
The name of the exported file.
options: dict
Display options, see `tf.advanced.options`.
!!! note "details"
* `condensed`
Has no effect. Exports to Excel will not be condensed, because the
number of columns is variable per row in that case.
Excel itself has nice possibilities for grouping rows.
You can also filter your tuples by means of hand-coding
before exporting them.
* `condenseType`
The condense type influences for which nodes
the full text will be exported.
Only nodes that are "smaller" than the condense type will have
their full text exported.
* `fmt`
This display parameter specifies the text format for any nodes
that trigger a text value to be exported.
* `tupleFeatures`
This is a display parameter that steers which features are exported
with each member of the tuples in the list.
If the iterable of tuples are the results of a query you have just
run, then an appropriate call to `displaySetup(tupleFeatures=...)`
has already been issued, so you can just say:
results = A.search(query)`
A.export(results)`
Results
-------
A file *toFile* in directory *toDir* with the following content:
There will be a row for each tuple.
If the input tuples are *uniform*, i.e. each tuple has the
same number of nodes, and nodes in the same column have the same node types,
then the result table has the following layout:
The columns are:
* **R** the sequence number of the result tuple in the result list
* **S1 S2 S3** the section as book, chapter, verse, in separate columns;
the section is the section of the first non book/chapter node in the tuple
* **NODEi TYPEi** the node and its type,
for each node **i** in the result tuple
* **TEXTi** the full text of node **i**,
if the node type admits a concise text representation;
the criterion is whether the node type has a type not bigger than the
default condense type, which is app specific.
If you pass an explicit `condenseType=`*xxx* as display parameter,
then this is the reference condenseType on which the decision is based.
* **XFi** the value of extra feature **XF** for node **i**,
where these features have been declared by a previous
displaySetup(tupleFeatures=...)`
If the input tuples are not uniform, the layout is more primitive.
There will be no header column, because the number of columns may vary per row.
A row contains the successive information of all nodes in a tuple.
Depending of the type of each node you get a number of columns of section information.
Then follow two columns with the node and the node type.
Depending on the type of the node, there follows a column with the text of the node.
No additional features are produced.
!!! caution "Encoding"
The exported file is written in the `utf_16_le` encoding.
This ensures that Excel can open it without hassle, even if there
are non-latin characters inside.
When you want to read the exported file programmatically,
open it with `encoding=utf_16`.
"""
display = app.display
if not display.check("table", options):
return ""
dContext = display.distill(options)
fmt = dContext.fmt
condenseType = dContext.condenseType
tupleFeatures = dContext.tupleFeatures
if toDir is None:
toDir = os.path.expanduser(DOWNLOADS)
if not os.path.exists(toDir):
os.makedirs(toDir, exist_ok=True)
toPath = f"{toDir}/{toFile}"
resultsX = getRowsX(app, tuples, tupleFeatures, condenseType, fmt=fmt)
with open(toPath, "w", encoding="utf_16_le") as fh:
fh.write(
"\ufeff"
+ "".join(
("\t".join("" if t is None else str(t) for t in tup) + "\n")
for tup in resultsX
)
)
# PLAIN and FRIENDS
def table(app, tuples, _asString=False, **options):
"""Plain displays of an iterable of tuples of nodes in a table.
The list is displayed as a compact markdown table.
Every row is prepended with the sequence number in the iterable,
and then displayed by `plainTuple`
!!! hint "condense, condenseType"
You can condense the list first to containers of `condenseType`,
before displaying the list.
Pass the display parameters `condense` and `condenseType`.
See `tf.advanced.options`.
Parameters
----------
tuples: iterable of tuples of integer
The integers are the nodes, together they form a table.
options: dict
Display options, see `tf.advanced.options`.
_asString: boolean, optional `False`
Whether to deliver the result as a HTML string or to display it directly
inside a notebook. When the TF-browser uses this function it needs the
HTML string.
"""
display = app.display
if not display.check("table", options):
return ""
api = app.api
F = api.F
fOtypev = F.otype.v
dContext = display.distill(options)
end = dContext.end
start = dContext.start
withPassage = dContext.withPassage
condensed = dContext.condensed
condenseType = dContext.condenseType
skipCols = dContext.skipCols
ltr = _getLtr(app, dContext) or "ltr"
if skipCols:
tuples = tuple(
tuple(x for (i, x) in enumerate(tup) if i + 1 not in skipCols)
for tup in tuples
)
item = condenseType if condensed else RESULT
if condensed:
tuples = condense(api, tuples, condenseType, multiple=True)
passageHead = f'</th><th class="tf {ltr}">p' if withPassage is True else ""
html = []
one = True
newOptions = display.consume(options, "skipCols")
for (i, tup) in tupleEnum(tuples, start, end, LIMIT_TABLE, item):
if one:
heads = '</th><th class="tf">'.join(fOtypev(n) for n in tup)
html.append(
f'<tr class="tf {ltr}">'
f'<th class="tf {ltr}">n{passageHead}</th>'
f'<th class="tf {ltr}">{heads}</th>'
f"</tr>"
)
one = False
html.append(
plainTuple(
app,
tup,
i,
item=item,
position=None,
opened=False,
_asString=True,
skipCols=set(),
**newOptions,
)
)
html = "<table>" + "\n".join(html) + "</table>"
if _asString:
return html
dh(html)
def plainTuple(
app, tup, seq, item=RESULT, position=None, opened=False, _asString=False, **options
):
"""Display the plain text of a tuple of nodes.
Displays the material that corresponds to a tuple of nodes
as a row of cells,
each displaying a member of the tuple by means of `plain`.
Parameters
----------
tup: iterable of integer
The members of the tuple can be arbitrary nodes.
seq: integer
an arbitrary number which will be displayed in the first cell.
This prepares the way for displaying query results, which come as
a sequence of tuples of nodes.
item: string, optional `result`
A name for the tuple: it could be a result, or a chapter, or a line.
position: integer, optional `None`
Which position counts as the focus position.
If *seq* equals *position*, the tuple is in focus.
The effect is to add the CSS class *focus* to the output HTML
for the row of this tuple.
opened: booolean, optional `False`
Whether this tuple should be expandable to a `pretty` display.
The normal output of this row will be wrapped in a
<details><summary>plain</summary>pretty</details>`
pattern, so that the user can click a triangle to switch between plain
and pretty display.
!!! caution
This option has only effect when used in the TF browser.
options: dict
Display options, see `tf.advanced.options`.
_asString: boolean, optional `False`
Whether to deliver the result as a HTML string or to display it directly
inside a notebook. When the TF-browser uses this function it needs the
HTML string.
Result
------
html string or `None`
Depending on *asString* above.
"""
display = app.display
if not display.check("plainTuple", options):
return ""
api = app.api
F = api.F
T = api.T
N = api.N
otypeRank = N.otypeRank
fOtypev = F.otype.v
_browse = app._browse
dContext = display.distill(options)
condenseType = dContext.condenseType
colorMap = dContext.colorMap
highlights = dContext.highlights
withPassage = dContext.withPassage
skipCols = dContext.skipCols
ltr = _getLtr(app, dContext) or "ltr"
if skipCols:
tup = tuple(x for (i, x) in enumerate(tup) if i + 1 not in skipCols)
if withPassage is True:
passageNode = _getRefMember(otypeRank, fOtypev, tup, dContext)
passageRef = (
""
if passageNode is None
else app._sectionLink(passageNode)
if _browse
else app.webLink(passageNode, _asString=True)
)
passageRef = f'<span class="tfsechead ltr">{passageRef}</span>'
else:
passageRef = ""
newOptions = display.consume(options, "withPassage")
newOptionsH = display.consume(options, "withPassage", "highlights")
highlights = getTupleHighlights(api, tup, highlights, colorMap, condenseType)
if _browse:
prettyRep = (
prettyTuple(app, tup, seq, withPassage=False, **newOptions)
if opened
else ""
)
current = "focus" if seq == position else ""
attOpen = "open " if opened else ""
tupSeq = ",".join(str(n) for n in tup)
if withPassage is True:
sparts = T.sectionFromNode(passageNode, fillup=True)
passageAtt = " ".join(
f'sec{i}="{sparts[i] if i < len(sparts) else ""}"' for i in range(3)
)
else:
passageAtt = ""
plainRep = "".join(
'<span class="col">'
+ mdEsc(
app.plain(
n,
_inTuple=True,
withPassage=_doPassage(dContext, i),
highlights=highlights,
**newOptionsH,
)
)
+ "</span>"
for (i, n) in enumerate(tup)
)
html = (
f'<details class="pretty dtrow {current}" seq="{seq}" {attOpen}>'
f"<summary>"
f'<a href="#" class="pq fa fa-solar-panel fa-xs"'
f' title="show in context" {passageAtt}></a>'
f'<a href="#" class="sq" tup="{tupSeq}">{seq}</a>'
f" {passageRef} {plainRep}"
f"</summary>"
f'<div class="pretty">{prettyRep}</div>'
f"</details>"
)
return html
html = [str(seq)]
if withPassage is True:
html.append(passageRef)
for (i, n) in enumerate(tup):
html.append(
app.plain(
n,
_inTuple=True,
_asString=True,
withPassage=_doPassage(dContext, i),
highlights=highlights,
**newOptionsH,
)
)
html = (
f'<tr class="tf {ltr}"><td class="tf {ltr}">'
+ f'</td><td class="tf {ltr}">'.join(html)
+ "</td></tr>"
)
if _asString:
return html
passageHead = f'</th><th class="tf {ltr}">p' if withPassage is True else ""
head = (
(
f'<tr class="tf {ltr}"><th class="tf {ltr}">n{passageHead}</th>'
f'<th class="tf {ltr}">'
)
+ f'</th><th class="tf {ltr}">'.join(fOtypev(n) for n in tup)
+ "</th></tr>"
)
html = "<table>" + head + "".join(html) + "</table>"
dh(html)
def plain(app, n, _inTuple=False, _asString=False, explain=False, **options):
"""Display the plain text of a node.
Displays the material that corresponds to a node in a compact way.
Nodes with little content will be represented by their text content,
nodes with large content will be represented by an identifying label.
Parameters
----------
n: integer
Node
options: dict
Display options, see `tf.advanced.options`.
_inTuple: boolean, optional `False`
Whether the result is meant too end up in a table cell produced by
`plainTuple`. In that case some extra node types count as big and will
not be displayed in full.
_asString: boolean, optional `False`
Whether to deliver the result as a HTML string or to display it directly
inside a notebook. When the TF-browser uses this function it needs the
HTML string.
explain: boolean, optional `False`
Whether to print a trace of which nodes have been visited and how these
calls have contributed to the end result.
Result
------
html string or `None`
Depending on *asString* above.
"""
return render(app, False, n, _inTuple, _asString, explain, **options)
# PRETTY and FRIENDS
def show(app, tuples, **options):
"""Displays an iterable of tuples of nodes.
The elements of the list are displayed by `A.prettyTuple()`.
!!! hint "condense, condenseType"
You can condense the list first to containers of `condenseType`,
before displaying the list.
Pass the display parameters `condense` and `condenseType`.
See `tf.advanced.options`.
Parameters
----------
tuples: iterable of tuples of integer
The integers are the nodes, together they form a table.
options: dict
Display options, see `tf.advanced.options`.
Result
------
html string or `None`
When used for the TF browser (`app._browse` is true), the result is returned
as HTML. Otherwise the result is directly displayed in a notebook.
"""
display = app.display
if not display.check("show", options):
return ""
dContext = display.distill(options)
end = dContext.end
start = dContext.start
condensed = dContext.condensed
condenseType = dContext.condenseType
skipCols = dContext.skipCols
if skipCols:
tuples = tuple(
tuple(x for (i, x) in enumerate(tup) if i + 1 not in skipCols)
for tup in tuples
)
api = app.api
F = api.F
item = condenseType if condensed else RESULT
if condensed:
tuples = condense(api, tuples, condenseType, multiple=True)
newOptions = display.consume(options, "skipCols")
for (i, tup) in tupleEnum(tuples, start, end, LIMIT_SHOW, item):
item = F.otype.v(tup[0]) if condensed and condenseType else RESULT
prettyTuple(app, tup, i, item=item, skipCols=set(), **newOptions)
def prettyTuple(app, tup, seq, item=RESULT, **options):
"""Displays the material that corresponds to a tuple of nodes in a graphical way.
The member nodes of the tuple will be collected into containers, which
will be displayed with `pretty()`, and the nodes of the tuple
will be highlighted in the containers.
Parameters
----------
tup: iterable of integer
The members of the tuple can be arbitrary nodes.
seq: integer
an arbitrary number which will be displayed in the heading.
This prepares the way for displaying query results, which come as
a sequence of tuples of nodes.
item: string, optional `result`
A name for the tuple: it could be a result, or a chapter, or a line.
options: dict
Display options, see `tf.advanced.options`.
Result
------
html string or `None`
When used for the TF browser (`app._browse` is true), the result is returned
as HTML. Otherwise the result is directly displayed in a notebook.
"""
display = app.display
if not display.check("prettyTuple", options):
return ""
dContext = display.distill(options)
colorMap = dContext.colorMap
highlights = dContext.highlights
condenseType = dContext.condenseType
condensed = dContext.condensed
skipCols = dContext.skipCols
_browse = app._browse
if skipCols:
tup = tuple(x for (i, x) in enumerate(tup) if i + 1 not in skipCols)
if len(tup) == 0:
if _browse:
return ""
else:
return
api = app.api
N = api.N
sortKey = N.sortKey
containers = {tup[0]} if condensed else condenseSet(api, tup, condenseType)
highlights = getTupleHighlights(api, tup, highlights, colorMap, condenseType)
if not _browse:
dh(f"<p><b>{item}</b> <i>{seq}</i></p>")
if _browse:
html = []
for t in sorted(containers, key=sortKey):
h = app.pretty(
t,
highlights=highlights,
**display.consume(options, "highlights"),
)
if _browse:
html.append(h)
if _browse:
return "".join(html)
def pretty(app, n, explain=False, **options):
"""Displays the material that corresponds to a node in a graphical way.
The internal structure of the nodes that are involved is also revealed.
In addition, extra features and their values are displayed with the nodes.
!!! hint "Controlling pretty displays"
The following `tf.advanced.options`
are particularly relevant to pretty displays:
* `condenseType`: the standard container to display nodes in;
* `full`: whether to display a reference to the material or the material itself;
* `extraFeatures`: additional features to display
* `tupleFeatures`: additional features to display (primarily for `export`.
Parameters
----------
n: integer
Node
options: dict
Display options, see `tf.advanced.options`.
explain: boolean, optional `False`
Whether to print a trace of which nodes have been visited and how these
calls have contributed to the end result.
Result
------
html string or `None`
When used for the TF browser (`app._browse` is true), the result is returned
as HTML. Otherwise the result is directly displayed in a notebook.
"""
return render(app, True, n, False, False, explain, **options)
def _getRefMember(otypeRank, fOtypev, tup, dContext):
minRank = None
minN = None
for n in tup:
nType = fOtypev(n)
rank = otypeRank[nType]
if minRank is None or rank < minRank:
minRank = rank
minN = n
if minRank == 0:
break
return (tup[0] if tup else None) if minN is None else minN
def _doPassage(dContext, i):
withPassage = dContext.withPassage
return withPassage is not True and withPassage and i + 1 in withPassage