From b8a5fbebb3e889d082823f5a5bd50db403c5ca13 Mon Sep 17 00:00:00 2001
From: IanCa <ianrcallanan@gmail.com>
Date: Thu, 7 Sep 2023 18:52:40 -0500
Subject: [PATCH 1/2] Add svg support for wordcloud

---
 hed/tools/visualization/tag_word_cloud.py     |  25 ++++++++++-
 hed/tools/visualization/word_cloud_util.py    |  41 ++++++++++++++++--
 tests/data/visualization/word_mask.png        | Bin 0 -> 4717 bytes
 tests/models/test_definition_dict.py          |  10 +++++
 .../visualization/test_tag_word_cloud.py      |  40 +++++++++++++++++
 5 files changed, 111 insertions(+), 5 deletions(-)
 create mode 100644 tests/data/visualization/word_mask.png

diff --git a/hed/tools/visualization/tag_word_cloud.py b/hed/tools/visualization/tag_word_cloud.py
index 68a3a257d..9f9092cba 100644
--- a/hed/tools/visualization/tag_word_cloud.py
+++ b/hed/tools/visualization/tag_word_cloud.py
@@ -1,9 +1,9 @@
 import numpy as np
 from PIL import Image
-from hed.tools.visualization.word_cloud_util import default_color_func, WordCloud
+from hed.tools.visualization.word_cloud_util import default_color_func, WordCloud, generate_contour_svg
 
 
-def create_wordcloud(word_dict, mask_path=None, background_color=None, width=400, height=200, **kwargs):
+def create_wordcloud(word_dict, mask_path=None, background_color=None, width=400, height=None, **kwargs):
     """Takes a word dict and returns a generated word cloud object
 
     Parameters:
@@ -25,6 +25,13 @@ def create_wordcloud(word_dict, mask_path=None, background_color=None, width=400
         mask_image = load_and_resize_mask(mask_path, width, height)
         width = mask_image.shape[1]
         height = mask_image.shape[0]
+    if height is None:
+        if width is None:
+            width = 400
+        height = width // 2
+    if width is None:
+        width = height * 2
+
     kwargs.setdefault('contour_width', 3)
     kwargs.setdefault('contour_color', 'black')
     kwargs.setdefault('prefer_horizontal', 0.75)
@@ -41,6 +48,20 @@ def create_wordcloud(word_dict, mask_path=None, background_color=None, width=400
     return wc
 
 
+def word_cloud_to_svg(wc):
+    """Takes word cloud and returns it as an SVG string.
+
+    Parameters:
+        wc(WordCloud): the word cloud object
+    Returns:
+        svg_string(str): The svg for the word cloud
+    """
+    svg_string = wc.to_svg()
+    svg_string = svg_string.replace("fill:", "fill:rgb")
+    svg_string = svg_string.replace("</svg>", generate_contour_svg(wc, wc.width, wc.height) + "</svg>")
+    return svg_string
+
+
 def summary_to_dict(summary, transform=np.log10, adjustment=5):
     """Converts a HedTagSummary json dict into the word cloud input format
 
diff --git a/hed/tools/visualization/word_cloud_util.py b/hed/tools/visualization/word_cloud_util.py
index 6071a138a..490be199f 100644
--- a/hed/tools/visualization/word_cloud_util.py
+++ b/hed/tools/visualization/word_cloud_util.py
@@ -7,14 +7,31 @@
 from wordcloud import WordCloud
 
 
-def _draw_contour(wc, img):
+def generate_contour_svg(wc, width, height):
+    """Generates an SVG contour mask based on a word cloud object and dimensions.
+
+    Parameters:
+        wc (WordCloud): The word cloud object.
+        width (int): SVG image width in pixels.
+        height (int): SVG image height in pixels.
+
+    Returns:
+        str: SVG point list for the contour mask, or empty string if not generated.
+    """
+    contour = _get_contour_mask(wc, width, height)
+    if contour is None:
+        return ""
+    return _numpy_to_svg(contour)
+
+
+def _get_contour_mask(wc, width, height):
     """Slightly tweaked copy of internal WorldCloud function to allow transparency"""
     if wc.mask is None or wc.contour_width == 0 or wc.contour_color is None:
-        return img
+        return None
 
     mask = wc._get_bolean_mask(wc.mask) * 255
     contour = Image.fromarray(mask.astype(np.uint8))
-    contour = contour.resize(img.size)
+    contour = contour.resize((width, height))
     contour = contour.filter(ImageFilter.FIND_EDGES)
     contour = np.array(contour)
 
@@ -22,6 +39,15 @@ def _draw_contour(wc, img):
     contour[[0, -1], :] = 0
     contour[:, [0, -1]] = 0
 
+    return contour
+
+
+def _draw_contour(wc, img):
+    """Slightly tweaked copy of internal WorldCloud function to allow transparency"""
+    contour = _get_contour_mask(wc, img.width, img.height)
+    if contour is None:
+        return img
+
     # use gaussian to change width, divide by 10 to give more resolution
     radius = wc.contour_width / 10
     contour = Image.fromarray(contour)
@@ -44,6 +70,15 @@ def _draw_contour(wc, img):
 WordCloud._draw_contour = _draw_contour
 
 
+def _numpy_to_svg(contour):
+    svg_elements = []
+    points = np.array(contour.nonzero()).T
+    for y, x in points:
+        svg_elements.append(f'<circle cx="{x}" cy="{y}" r="1" stroke="black" fill="black" />')
+
+    return '\n'.join(svg_elements)
+
+
 def random_color_darker(word=None, font_size=None, position=None, orientation=None, font_path=None, random_state=None):
     """Random color generation func"""
     if random_state is None:
diff --git a/tests/data/visualization/word_mask.png b/tests/data/visualization/word_mask.png
new file mode 100644
index 0000000000000000000000000000000000000000..e235d063ea190f92c537137ad9eb3a87b3faf3e4
GIT binary patch
literal 4717
zcmcIodpOkF*I#2WGf87iLS-7`)^YFRRwiSNTZokCRGQp7IU#jolrv?75-N2_$>=D_
z5l1048l@=7sH4cGDP1O%OViMi{PxK2yyuViulIRg&*Pc>U3-7mT6^u!TI=)OX+CRQ
zRcFng1pun<ZY*B_1rGY3gI7ip3+;j~^rINdbYG80fBWzusQ~Kw?kvakac4gCbnWVL
zQ)f6v-=3Ok?!ax@;cn=>`EOspvtO!6<jRHWl2v7flE%8B31wi|O|>Lq0HnH7fWzv@
zgfoE-+(I4&7;BkCC|JzLflsyw3yW7vFfiBP|9#o5uv@_1lG)hTJk>k4|FH?C*j<9T
zxG%GkG_@0hk3K5G=J#vw4&M@Q_DYQ*!JK0qzHx&yS&et2lKHsoSvPwh)%o96%W;>~
z-&PF#wY4{R1eZ>eve(k0D8#dH9#0QQ7anPDZGk@sbgj}o5jsaP2Yzd93^yF-UZ&j8
zP&`fcyBc;$xy-UH=}5yuz1)u&e>~mRUA*K;--xYFf?3(d0|Yy}+GK2lj(xbwfeufp
zSSbcK6z{7QMCJ@16($X=df`w=e2X4Zgh^GN6YN#|#>Wjh>3#RNX?aU3^bg=Z+4d&K
z@F>k{d(|=}Np(ER8-jgPeVh4JBk5ufvE!P9#OEYVw%%}oay@R0K!2bX=W{}wG{B==
zN?RF_g0m6doI*jWc{LK3Y9Ej@u_$9no1Uu5%HdqSEpgrwQoazI_{vQ)kj&v3409_z
zzlcrvuC@xq8js=nC}yS)iy}H6yk&!t6z?d;2y#a&X#B?Y3Pdfh;R<zr<6H$|PD_Qk
z$#ByB*+!$>h_CpSez}q**~xQL2Pfu-K6d@ZkTY9vlpE>!che(Qrmkr=Bdo5~bptcA
zOp)SqRH8?At=^61`%7b4;f1c6^BmY3hWl7yg|4wQV!+AInGrmS*T&b}_ITgl+<26h
zfDZVkzK=zfM%)fk9zOH=ST2uJy|Doj!)hB%^K0LztIQL|FWPbOVyyOER_3~|m*$FB
z4J<yNBN@><LSihskdV$UY+Pr!7lH$Jymyz3Bnz=xt+gH&89O_xT4-3yyz`qmmTgY2
zF^WN67YxXP<c*#Xw6y2E23g>oX3rfjZSOXa&I@@7O?OVW=rE&n^NW?5?mutwL+Do@
zPT5nY#!)G^g}=7Z-_7>!nHNHW;twNnakSr@c!9bKM7B=i!DPLKJ@a%&;c3X^J0nBR
z{p<A@uD_>Log5F5d6V2{yg+7tp^|vzghDG`z*3}uLxt<`S&_%>78M~$rm~96A=GTg
z*xsqD$MH4K<Y{nC1{X!<o%1I&tImS%&_d0uCTxMy1c7OxWn%zlo&HY^r3x|26hujr
zp<9?1a4QqXh?3=65d`<U;~+h#bz`O|uSpdrjF9DD!wcNIBY+Xbj8Zs;6MD+{d+>t#
z?#;krJN%d=TF~UB#L$-&{mtv8?lc7Ua=s0>2}?CjSpPHg$}$@@XzRS(ug=lA9s^Cc
z^ZibU1{zec@X%1YLM<LbLZ(VwC01_@F;J++0pXN(z+W`!{9UmC$M6Cd76tq|p7g76
zRBLpBoBF+^3>iML2dXiDZxY&5Nn_u(b0#Q@!6Hr2&FghG-v=AEwKjVoNtE4#7I(=O
zF%!r3lL<HTdfm;VVMDmZLnkzO5luQu&WaBzbFV0FjYLOP<286!UkY49&#_k|4(@t1
zMy!<(wvaglnx=3jcqrXbB6!JE;_evae<BDp_q76ito-effiypQtpa1EtUaFBYh&&L
z5v|`=q=+UOobfQ(5#C7V{8}>(1r8G*0+>;;Z5W}OtS9Cak*>J_(!-~8g1?~3_*N}p
za`N<~LRd@;2HJCxdr!1(O0R`Tt``G!&crT<=H&b%1b9{9Fjlc%ay=EeCr!y58WDmj
z{i~~4lIzhx#U=c*>5WsX5*A8Jc1zOip{em(&zeqA>@i>j%0{eLAId+1g@OpA{EK#X
zw}1ao?H)&fuc3uF;Dx5fX38sciKpa;L@2&b`E^S!dPZrFCc{hidZmdzy~M&-lMk(z
zQ&$4_T+tc13GSHhAqtdJlJY-IpUr!wcMDcnyIW{?iP@T>0@0gG<XUY1+W~w>{L1=#
zzq0=%&IPeOS>7g?-crv|fU;$6Sa`f-_6ir|-&_rAY4!>b(siA@Ut^$Kb1CW>HQv)R
z=F){-3UHNVEJfg8tlY+wuehpVh*mtZ)aO{RBwq@EYqGqdm5E#u`UMSLd(VG&2Lm4;
zlUu!r$mG0+-dh(A<r5Vl!hO>|D>LanHDD0u_gT1Bqh=N~WFz<XuWZq}PK0^R9TpK5
zM}F>DK;c}z^%8^uCe<lv6!7USsW3udF(}nakT<zvKEFE19Sbp~PiclTJ&^r_iPz`v
zm?c8MBja^+q}g<PatHa}O|rbSzp>iMDeA@8v8lZP9`Zn(+7MrYw8P=~6i!AI2JTtP
zh5nz%H^^gPWguPnLII@V=;X@V?dvrDtInZh#@;&Sl^@NpuxEY$-02`6N(n4ltqebv
zJT`H1y7C}T8L|$X-vnGE<JrZBk^_@%R3N}ZmZ+`H>3pFN-1LP#3h<TsNm!B~2~9+~
zxK!>`oZf(Vx`Ve{N&kFm0m8l7H}0Y6OgH#wpQ=mw(+Jum2{j1>uHI6s3>2qn3LyU7
zbb9a8ETD#ZXZ`e1%87va;Z^PbL}LNX`=4-qXASGCMO|ZpXSe19HM{u56S)`PNd$61
z9)gF#C<G!#!_fO_%2At=--mW(=Q$lZb<PD8m_F4!Ng6u!ibIs1!*nS4Gs3%f&QwNz
z2Ntw~AKcLjLL0cX-Mi=Gg!wqYiqqdd86PA<vHs&R>4lVDTs_jiAxAVpl>t1*Ya)eL
z<veHpkx{z*H2;U!>{fgE{dA`J%`9zBq6ZPgby;YUaK9WOCpw8CJXLyQHNY79Ma6?8
zKeyyL)1qg|a|Az=69D{^5k{fRoiy$JyCVE|ad;8hAv|)|m=-q{&H65OEmD~o{672p
zVCZ+%?*j*S9i;zQ3uSc?QLz8T#an3rBs3(V9g8n7<yZeKA5{9z44kfux~wY@pni6C
zit4ie9Rohr(p9N+VEes60?#qvTA7hJiyBbAA1`7?b)kW!3-4`!xwNWftvQq*GH@D7
zLQyr|v;s+XqZsb^G95SWPrP0uWUH7Z+26NV6KYIWd*B6{h5Hz2w*<kWoR1jDB?z#J
ziy-TmXr<#Idc+G($Ss<pf_{()scR%X`ZC}a^8&F*(fHs<+`~jdR~11Jw(yMtwBZH!
zkT+_{q5%eanUoQg94I%K<0aYiIw1jsp3EpfTYE^ADlh55g3E-}gi`6-ghA)=LPPL5
zA<FX{$AADYV1HaVbeIL=WqkkC_f#0p6`7zJ;d5RuGf+oGg#n02j>r-YFx6m}I_Ksa
zycX;lO}c;PZOS<e756VV#H1k{rv8I+#ELh+9iIn?-u*A^xY3N#1l)JD!DGysWJ0Lw
ztHPh(4&tqHQN>~>HNGo(t$759E&gStEnRNK8I^?PbGHxcN+XOP-&KXnawA7s3*zwK
zS$K={#FN=jD$m^IYB3^Rsf70G?i{#0jXrtW<Q3+J5a_jOoHne`7?mrb>M`5`#KmY^
zzk<(M(fO4XIEG)_n7oKbJ4&0ntoe;p?VZYCYusFCDAh7-_QHUl+!%SOsyUD16k2Ns
z;v10@Iild2NE-q~ZhAE7Az9*lFHZ^nByl|TWY9z#$TyJY7!Mr93Rf&0@?&PcU;2Up
zf1TQ&sm|eLO}Z!)<j+btA<|2qWiuC8G%4PB5ToXJa?(vgT5~c_g?q;MgA+3|OSa!)
z&v{Ii(=~ysg!JRJ1#`d!y>BD~E&i)SJ*itg0w&$BCm8YG^#T$m=i+EDP3m@WRFwc`
z1iJ68Y!!w|`<Oc4^{0bG?&S>~4m^rMX>_H6a3eEwDFUGFPbVjtlC_VFZ7kqf=~Zt$
zef82FtT?w_rp$NUez`{^$H;m6Cv>R#xy9kD3jLJsI0PIO1=}4Yvgw!0CYYI_e{h~)
zOT*rL4wO_Ze?sRz->|hWnG$)BSU?;2yGq{lsmEK7-0mC68DITW78xUphcjaxZ_K~C
zvD0Hhei+>duJ`j^pihLOnRl5XS3^n4oQgVm(>5a14g6?o)BCH~=+C#PPiqn~#yDmc
zy$ND%_fM#oMYoM9g?=6{7b+~<n3TV)Eon_j%8S9G^M6mMEVN0WANG{gQ#)_#-+7$%
zfv)ttxqL&*Ysoz2_-k4%85&3Qvrc{<pp)s;Yn}Yr4|TU@jQAH|E^Qp#X7=sDgIIZc
zc$s!%2al4~TQ`&u&8NofJtxAhR!hh_SH1(mJ~Wk{(|3`Pu*dsw?Ggm3=%$ml<lV6Z
zdcoXv^cwFo?JwW?ad$=f!Yw4yufRR4+4~cT!`NtT%HA`R1leIFxW?;S`;j=_+9IxX
zzwZeV_7t0(^sIZ)9s-@MZM}?b#UutC?nt`QO}>UCcGlKx4J0|<fgP)##iNv)<}Ead
zoxI{G!FW`5R?wbJu0TTjJc`dr5w@d&f)XD)?lg#L$J4pq4V3ol_ny(DP**|nZL1q9
z`=CUibB!WX8A?o|Sg8URg8FeHRaPwAbJHAULUP(oWRq(XnH}=_rAn2T1jDdX$Rwh*
zu)qkFF)Ww)mCtMCQG|}xK83MIN6(6|w)HNn*Oz(aqs+@X)6}QI$3-9aI_b7K^D7Iz
z?n}$ui<T={2ZY6{?&2X0AuNh3vT-D`D*f#ht3K12)UT=EoxFDQj?JhKQEiVkqOT0G
zVV))ZdR-0TwaeMdE?5x*MF-LQY(>f+Xm2i&kFJVW6c%l*+zLP)mLNt`pqppqU$awk
z7qxFgQk)dAr}<rD`xERbm0Oh^(^k6d+N}afsrD|YmgsbdzC4V2wd9BhD`~iwAq=*D
z(T3%^uyopx?hN~!wRz=wEBH8}X6vgrYf5Dqe2T7_xcF7YjnkOg=H=~`W(*z5J(%v>
zv)QYaQ4t1ZQi3pI=A13FE>SFx%b5eJP_xmYqfc!Zr#*@KhH*gF$vOQPy_rfT)W5(%
zUm-*`tpX=SOnv&c;Y4mfMt~N4dug<xK$(wI<6feOl4kTNZY#lEl~Y6cy~F|5>Hc^s
zc66o)bJ;Yn&)jZ=PB};6{<hDXxC>2T?CO+$M4uw^TMTZ2=V$JXnHAJ>{5m=@gb_I7
zxY}mkB}x{;(6s0JmI)>A5;*c_x1e7JF<)kUX3#EhgxWK&WTjuCM5CJ!v6^=;a}|?F
z)m=vy<N3I(><(=2Og-i{$oAWGVgqL`Qn(7pg>W8)K?)*{uaRKp(qvqOD#j8)kT*-N
zqoCmFTEkJQ5g+H1ZHq-S+sqQfy)Bu@O*Y0q$@{hAPWgQZ-1~yBZtNS8ox{ijprRgI
iu8jEZ|KYA?3Y(5OowLqx@ojW}1n$mjSY=H1!T$jA_mTYo

literal 0
HcmV?d00001

diff --git a/tests/models/test_definition_dict.py b/tests/models/test_definition_dict.py
index 357584cc1..5005f55c5 100644
--- a/tests/models/test_definition_dict.py
+++ b/tests/models/test_definition_dict.py
@@ -134,5 +134,15 @@ def test_expand_defs(self):
             hed_string.expand_defs()
             self.assertEqual(str(hed_string), expected_results[key])
 
+    def test_altering_definition_contents(self):
+        def_dict = DefinitionDict("(Definition/DefName, (Event, Action))", self.hed_schema)
+        hed_string1 = HedString("Def/DefName", self.hed_schema, def_dict)
+        hed_string2 = HedString("Def/DefName", self.hed_schema, def_dict)
+        hed_string1.expand_defs()
+        hed_string2.expand_defs()
+        hed_string1.remove([hed_string1.get_all_tags()[2]])
+
+        self.assertNotEqual(hed_string1, hed_string2)
+
 if __name__ == '__main__':
     unittest.main()
diff --git a/tests/tools/visualization/test_tag_word_cloud.py b/tests/tools/visualization/test_tag_word_cloud.py
index 2b515c941..6bb940eec 100644
--- a/tests/tools/visualization/test_tag_word_cloud.py
+++ b/tests/tools/visualization/test_tag_word_cloud.py
@@ -2,12 +2,19 @@
 from wordcloud import WordCloud
 from hed.tools.visualization import tag_word_cloud
 from hed.tools.visualization.tag_word_cloud import load_and_resize_mask
+from hed.tools.visualization.word_cloud_util import generate_contour_svg
+
 import numpy as np
 from PIL import Image, ImageDraw
 import os
 
 
 class TestWordCloudFunctions(unittest.TestCase):
+    @classmethod
+    def setUpClass(cls):
+        cls.mask_path = os.path.realpath(os.path.join(os.path.dirname(__file__),
+                                                      '../../data/visualization/word_mask.png'))
+
     def test_convert_summary_to_word_dict(self):
         # Assume we have a valid summary_json
         summary_json = {
@@ -40,6 +47,30 @@ def test_create_wordcloud(self):
         self.assertEqual(wc.width, width)
         self.assertEqual(wc.height, height)
 
+    def test_create_wordcloud_default_params(self):
+        word_dict = {'tag1': 5, 'tag2': 3, 'tag3': 7}
+        wc = tag_word_cloud.create_wordcloud(word_dict)
+
+        self.assertIsInstance(wc, WordCloud)
+        self.assertEqual(wc.width, 400)
+        self.assertEqual(wc.height, 200)
+
+    def test_mask_scaling(self):
+        word_dict = {'tag1': 5, 'tag2': 3, 'tag3': 7}
+        wc = tag_word_cloud.create_wordcloud(word_dict, self.mask_path, width=300, height=300)
+
+        self.assertIsInstance(wc, WordCloud)
+        self.assertEqual(wc.width, 300)
+        self.assertEqual(wc.height, 300)
+
+    def test_mask_scaling2(self):
+        word_dict = {'tag1': 5, 'tag2': 3, 'tag3': 7}
+        wc = tag_word_cloud.create_wordcloud(word_dict, self.mask_path, width=300, height=None)
+
+        self.assertIsInstance(wc, WordCloud)
+        self.assertEqual(wc.width, 300)
+        self.assertLess(wc.height, 300)
+
     def test_create_wordcloud_with_empty_dict(self):
         # Test creation of word cloud with an empty dictionary
         word_dict = {}
@@ -54,6 +85,15 @@ def test_create_wordcloud_with_single_word(self):
         # Check that the single word is in the word cloud
         self.assertIn('single_word', wc.words_)
 
+    def test_valid_word_cloud(self):
+        word_dict = {'tag1': 5, 'tag2': 3, 'tag3': 7}
+        wc = tag_word_cloud.create_wordcloud(word_dict, mask_path=self.mask_path, width=400, height=None)
+        svg_output = tag_word_cloud.word_cloud_to_svg(wc)
+        self.assertTrue(svg_output.startswith('<svg'))
+        self.assertIn("<circle cx=", svg_output)
+        self.assertTrue(svg_output.endswith('</svg>'))
+        self.assertIn('fill:rgb', svg_output)
+
 
 class TestLoadAndResizeMask(unittest.TestCase):
     @classmethod

From 94ec01376159649008901612808d8e09697b46f5 Mon Sep 17 00:00:00 2001
From: IanCa <ianrcallanan@gmail.com>
Date: Thu, 7 Sep 2023 19:00:40 -0500
Subject: [PATCH 2/2] Add another function to init

---
 hed/tools/visualization/__init__.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/hed/tools/visualization/__init__.py b/hed/tools/visualization/__init__.py
index a40c0333b..389ba92f8 100644
--- a/hed/tools/visualization/__init__.py
+++ b/hed/tools/visualization/__init__.py
@@ -1 +1 @@
-from .tag_word_cloud import create_wordcloud, summary_to_dict
+from .tag_word_cloud import create_wordcloud, summary_to_dict, word_cloud_to_svg