langsci-forest-setup.sty

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%%      File: langsci-forest-setup.sty
%%    Author: Language Science Press (http://langsci-press.org)
%%      Date: 2016-01-16 16:47:43 UTC
%%   Purpose: This file contains optional styles and settings for the
%%            forest package. This version is for 2.0.
%%  Language: LaTeX
%%   Licence:
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

% use lualatex for subnodes in trees. St. Mü. 2025-04-07

% http://en.wikibooks.org/wiki/LaTeX/Linguistics#Syntactic_trees
\usepackage{tikz-dependency}

% dangerous, use at own risk
%\usetikzlibrary{tikzmark} % saves positions in tikz pictures, allows to connect arbitrary text


% http://tex.stackexchange.com/questions/256280/drawing-complex-dependency-diagrams-with-tikz-forest
\usetikzlibrary{arrows.meta}

\tikzset{deparrow/.style={-Latex}}%,blue}}


% externalization works with package memoize
\usepackage[linguistics]{forest}
%\usepackage{morewrites}



% specification for all trees, "default preamble" appends to existing specification.
% The version with apostrophe replaces it.
\forestset{default preamble'={
    for tree={align=center,parent anchor=south, child anchor=north,base=bottom,anchor=north},   
% anchor=north was introduced to get alignments of AVMs right.
% used to be anchor=north but this fails on trees in \ea \z examples, St. Mü. 02.05.2020 or may be not
%
% Saso 14.11.2020: align=center puts nodes into a tabular environment and adds spece between nodes.
% Me: If you want to use it somewhere, eg for glosses, put it in a style and say so explicitely.
%
% This would align trees to the baseline. We do not want this for TAG
% where several trees have to be aligned with respect to their center.
%    before drawing tree={
%      sort by=y,
%      for min={tree}{baseline}
%    }
  }}


\forestset{
  declare dimen={child anchor yshift}{0pt},
  adjust childrens child anchors/.style={
    if n children>=2{
      before packing={
        tempdima/.max={max_y}{children},
        for children={
          child anchor yshift=tempdima()-max_y()
        },
      }
    }{}
  },
  default preamble={
    for tree={
      edge path'={(!u.parent anchor)--([yshift=\forestoption{child anchor yshift}].child anchor)},
      adjust childrens child anchors
    }
  },
}



% http://www.texample.net/tikz/examples/borrowers-and-lenders/
\usetikzlibrary{arrows} %,positioning}
\tikzset{
    %Define standard arrow tip
    >=stealth'}



%% \forestset{
%% sn edges/.style={for tree={parent anchor=south, child anchor=north,align=center,base=bottom,where n children=0{tier=word}{}}},
%% background tree/.style={for tree={text opacity=0.2,draw opacity=0.2,edge={draw opacity=0.2}}}
%% }



% The following code fixes the size of triangles for examples with translations. The specification
% ",delay=with translation" at the leaves is important to make this work.
% http://tex.stackexchange.com/questions/167978/smaller-roofs-for-forest/205311#205311
\forestset{
  with translation/.style={
    l sep=0,inner xsep=0,
    append translation/.expanded/.wrap pgfmath arg={\gettranslation{##1}}{content},
    content/.expanded/.wrap pgfmath arg={\gettext{##1}}{content},
  },
  append translation/.style={
    if={strequal("#1","")}{}{
      append={[#1,no edge,inner xsep=0,inner ysep=0,outer ysep=0,
               tier=gloss, % does not work for TAG trees since the words are not aligned at the baseline
               l=\baselineskip + 0.2ex,
        ]}
    }
  }
}

\forestset{
  with unaligned translation/.style={
    l sep=0,inner xsep=0,
    append unaligned translation/.expanded/.wrap pgfmath arg={\gettranslation{##1}}{content},
    content/.expanded/.wrap pgfmath arg={\gettext{##1}}{content},
  },
  append unaligned translation/.style={
    if={strequal("#1","")}{}{
      append={[#1,no edge,inner xsep=0,inner ysep=0,outer ysep=0,
               %tier=gloss, % does not work for TAG trees since the words are not aligned at the baseline
               l=\baselineskip + 0.2ex,
        ]}
    }
  }
}


\def\gettext#1{\gettextA#1;;\endget}
\def\gettextA#1;#2;{\removesep#1;}
\def\gettranslation#1{\gettranslationA#1;;\endget}
\def\gettranslationA#1;#2;{\removesep#2;}
\def\removesep#1;#2\endget{#1}

% copy of the above for TAG
\forestset{
  with translation TAG/.style={
    l sep=0,inner xsep=0,
    append translation TAG/.expanded/.wrap pgfmath arg={\gettranslation{##1}}{content},
    content/.expanded/.wrap pgfmath arg={\gettext{##1}}{content},
  },
  append translation TAG/.style={
    if={strequal("#1","")}{}{
      append={[#1,no edge,inner xsep=0,inner ysep=0,outer ysep=0,
               %tier=gloss, % does not work for TAG trees since the words are not aligned at the baseline
               l=\baselineskip + 0.2ex,
        ]}
    }
  }
}


% a specific style that specifies the word tier: all nodes that do not have any children
% to do: make this style default and change all other figures explicitely as is described here:
% http://tex.stackexchange.com/questions/167983/getting-rid-of-a-default-where-specification
\forestset{
sm edges/.style={for tree={parent anchor=south, child anchor=north,base=top}, % changed base=bottom
                                % to top since we have NP and AVM nodes and the NP has to start
                                % right away at the same hight as the AVM St. Mü. 18.10.2017
                 where n children=0{tier=word,delay=with translation}{} %,delay=with translation
                 },
sm edges without translation/.style={for tree={parent anchor=south, child anchor=north,base=bottom},
                 where n children=0{tier=word}{}
                 },
word tier/.style={where n children=0{tier=word}{}
                 },
%
% dependency grammar trees, projective trees only, I guess. St. Mü. 29.09.2014
dg edges/.style={for tree={parent anchor=south, child anchor=north,align=center,base=bottom},
                 where n children=0{tier=word,edge=dotted,calign with current edge,delay=with translation}{}
                },
%
% Adjunct edges as suggested by Engels, 1977
% http://tex.stackexchange.com/questions/204770/defining-a-special-edge-style-for-connecting-nodes-in-forest/204819#204819
dg adjunct/.style={edge path={\noexpand\path[\forestoption{edge}]
(!u.parent anchor)--(.child anchor)-- +(0,#1)\forestoption{edge label};}},
%
dg adjunct/.default=6pt,
%
% for Tesniere-style translations
dg transfer/.style={edge path={\noexpand\path[\forestoption{edge}, rounded corners=3pt]
    % the line downwards
    (!u.parent anchor)-- +($(0,-l)-(0,4pt)$)-- +($(12pt,-l)-(0,4pt)$)
    % the horizontal line
    ($(!p.north west)+(0,l)-(0,20pt)$)--($(.north east)+(0,l)-(0,20pt)$)\forestoption{edge label};},!p.edge'={}},
% for Tesniere-style junctions
dg junction/.style={no edge, tikz+={\draw (!p.east)--(!.west) (.east)--(!n.west);}    },
%
%
%
% http://tex.stackexchange.com/questions/256280/drawing-complex-dependency-diagrams-with-tikz-forest/256300#256300
% in the tree draw the highest nodes of the inheritance network. For lower nodes insert empty nodes
% "[]". These will be filled by the same word and the numbers will be provided with bars according
% to remoteness from the word level.
%
% So cool! Thanks Sašo Živanović
%
% modified to cope with multiple occurances of the same word (the, the1, the2)
%
% If you want to put it into the style, put ",font=\it" after "content format={...}" 
%
  wg/.style={
    for tree={
      no edge,
      draw,
      outer ysep=1pt,
    },
    delay={
      copy label,
      for tree={nameN/.option=content},
    },
    for tree={content format={\strut\forestoption{content}}},
    where n children={0}{
      tier=word,
    }{},
  },
  copy label/.style={
    for children={
      if content={}{
        % content/.pgfmath={content("!u")},
        content/.option=!u.content, % faster
        calign with current,
        edge={draw,{-Triangle[open,reversed]}},
        copy label,
        % !u.content/.pgfmath={content},
        !u.content/.option=content, % faster
        !u.content+=',
      }{
        copy label,
      }
    }
  },
  % 1. This is not effective if there's many clashes.
  % 2. Register tempcounta must be free to use.
  nameN/.style={
    if nodewalk valid={name=#1}{
      tempcounta=2,
      @nameN=#1,
    }{
      name=#1,
    },
  },
  @nameN/.style={
    while nodewalk valid={name/.process=Rw1{tempcounta}{#1##1}}{
      tempcounta+=1,
    },
    name/.process=Rw1{tempcounta}{#1##1},
  },
%
%
% TAG
% http://tex.stackexchange.com/questions/207459/getting-old-school-trees-with-forest
% fit -> fit=band causes all nodes to have nothing below them
tag/.style={for tree={parent anchor=south, child anchor=north,align=center,base=top,fit=rectangle},
  where n children=0{delay=with translation TAG}{},
% the following breaks the alignments in sets
  begin draw/.code={\begin{tikzpicture}[baseline=(current bounding box.center)]}
},
%
% like the tag style, but without fit rectangle
no word baseline/.style={for tree={parent anchor=south, child anchor=north,align=center,base=top},
  where n children=0{delay=with translation}{}},
%
% a style that creates an arrow pointing to the substitution node from a tree top node encoded as daughter
substitution/.style={edge={<-,dashed},l+=\baselineskip},
%
%
% auxiliary nodes without node label
%empty nodes/.style={
%  delay={where content={}{shape=coordinate,for parent={for children={anchor=north}}}{}}}
}

% draws a red box about the area that spans the current node,
% the first daughter node and the last daughter node.
% label adds a label to the box. You can use schema=3 to give a certain local tree a
% number attached to the box.
\forestset{
  schema/.style={
    tikz={\node [draw,red,inner sep=-2pt,fit=()(!1)(!l),
              label={[font=\scriptsize,inner sep=0]south east:#1},
              ] {};},
    % the mother gets padding to the west and to the east.
    padding west=0.3333em, padding east=0.3333em,
    % the first child gets padding to the west
    for 1={padding west=0.3333em},
    % the last child gets padding to the east
    for -1={padding east=0.3333em},
    }
    % % the mother gets padding to the west and to the east.
    % padding west=1em, padding east=1em,
    % % the first child gets padding to the west
    % for 1={padding west=1em},
    % % the last child gets padding to the east
    % for -1={padding east=1em},
    % }
}


% Manual page 52
%% \forestset{
%% nice empty nodes/.style={
%% for tree={calign=fixed edge angles},
%% delay={where content={}{shape=coordinate,for parent={for children={anchor=north}}}{}}
%% }}


% forest 2.0
\forestset{
     empty nodes/.style={
     delay={where content={}{shape=coordinate,for siblings={anchor=north}}{}}
     }
   }

\forestset{
  % Easy extra edges
%  edge to'/.style 2 args={
%    tikz+={\path[#2](#1.parent anchor)--(.child anchor);}
%  },
  edge to'/.style 2 args={
   tikz+={\path[#2](#1.parent anchor)--([yshift=\forestoption{child anchor yshift}].child anchor);},
  }, 
  edge to/.style={
    edge to'/.expanded={\unexpanded{#1}}{\forestoption{edge}},
  },
%  edge from'/.style 2 args={
%    tikz+={\path[#2](.parent anchor)--(#1.child anchor);}
%  },
  edge from'/.style 2 args={
   tikz+={\path[#2](.parent anchor)--(\forestoption{child anchor yshift}]#1.child anchor);}
  }, 
  edge from/.style={
    edge from'/.process=_O{#1}{#1.edge},
  },
  % Identify two (adjacent, vertically aligned) nodes.  See the explanation of
  % how this works in "type-hierarchies.tex"
  identify/.style={
    phantom,
    temptoksa/.option=#1.name,
    before typesetting nodes/.process=Rw1{temptoksa}{
      content/.option=##1.content,
      content format/.option=##1.content format,
      node format/.option=##1.node format,
    },
    for parent/.process=Ow1{n}{
      after packing node={
        for ##1={
          ignore,ignore edge,forget@node@boundary/.option=grow,
        },
      },
    },
    for nodewalk/.process=Rw1Ow1{temptoksa}{name=##1}{!u.name}{edge to={##1}},
  },
  forget@node@boundary/.code={%
    \forestolet{positive@edge@#1}\relax
    \forestolet{negative@edge@#1}\relax
  },
  % Uniformly spread any sequence of nodes
  spread/.style n args=4{
    % #1 = coordinate (x,y,s,l)
    % #2 = nodewalk to spread
    % #3 = relative node name of the node with the final coordinate
    % #4 = final coordinate shift
    for nodewalk={
      tempcounta'=-1,
      #2,
      tempdima/.pgfmath={(#1("#3")+#4-#1(""))/tempcounta()}
    }{tempcounta'+=1},
    tempdimb'=0pt,
    for nodewalk={#2}{
      #1'+/.register=tempdimb,
      tempdimb'+/.register=tempdima,
    },
  },
}

% The old type hierarchy style
\forestset{
typehierarchy/.style={for tree={parent anchor=south, child
% old version:
%    anchor=north,align=center,base=top,font=\itshape,fit=rectangle}},
% to get parallel lines:
    anchor=north,align=center,base=top,font=\itshape,calign=fixed angles}},
% if there are instances as leaves in the hiarchy
instance/.style={edge=dotted}
}

% Use
% type hierarchy=\vphantom{d} 
% if needed to pass the phantom command to the style
% It is then included into every content cell.
% Saso 29.07.2024

% Type hierarchy (Saso 2020)
\forestset{
  type hierarchy/.style={
    for tree={
      anchor=base, % does not work for AVMs in hierarchies St. Mü. 10.02.2021
      %anchor=north, % does not work either since "case" would be higher than "vform" I fixed this
                    % with \vphantom{vform} next to the "case" type. St. Mü. 27.07.2021
                    % but this would have to be done for all type hierarchies (e.g. also for simple
                    % ones with noun and verb, so, I stick to anchor=base and fix it for hierarchies
                    % with AVMs seperately.
      content format={\noexpand\type{\unexpanded{#1}\forestoption{content}}},
    },
    % The \ldots should not be in italics, so we add \normalfont to the original definition. Saso 27.05.2020
    TeX={
      \let\origldots\ldots
      \def\ldots{\normalfont\origldots}%
    },
    % Saso suggested the above rather than looking at the content, but we may use looking at the
    % content for replacing "...":
    delay={where content={...}{content=\normalfont\ldots}{}},
  },
% https://tex.stackexchange.com/a/722485/18561
  auto name/.style={
    before typesetting nodes={
    for tree={alias/.option=content}, % give the nodes the name of the type, so that we can
                             % draw additional lines with \draw instead of doing intransparent
                             % node walks. 
    }},
  attrib/.style={
    content format={\noexpand\attrib{\forestoption{content}}},
    draw,
  },
  % Partitions are typeset in small caps (\textsc) and put in boxes (draw).
  partition/.style={
    content format={\noexpand\textsc{\forestoption{content}}},
    draw,
  },
  instance/.style={
    gray,edge={gray,dashed},align=center
  },
  % if all leaves in the hierarchy are instances, we can declare this as follows:
  % nodes which do not have children (n children=0) are drawn in gray and edges leading to them are dashed 
  instances/.style={
      for tree={
      s sep=0,
      if n children=0{
        instance   % easier to maintain if we use the style than repeat the definition of the style
      }{},
  }},
}

% to add lines in hierarchies that leave in the south and go into north
% can be used in connection with "auto name" for types and then \draw (type1) to (type2);
% Example picture of Haugereid's hierarchy in GT book. 24.05.2025
\tikzset{from south line to north/.style={to path=(\tikztostart.south)--(\tikztotarget.north)\tikztonodes}}