From 205403d377d4b6cb7f2b0fb48044442069269992 Mon Sep 17 00:00:00 2001
From: Claude Paroz <claude@2xlibre.net>
Date: Tue, 22 Oct 2019 20:45:21 +0200
Subject: [PATCH] Fixes #202 - Keep error content when importing xls files

---
 HISTORY.md                 |   1 +
 src/tablib/formats/_xls.py |   5 ++++-
 tests/files/errors.xls     | Bin 0 -> 5632 bytes
 tests/test_tablib.py       |  29 ++++++++++++++++++++++++-----
 4 files changed, 29 insertions(+), 6 deletions(-)
 create mode 100644 tests/files/errors.xls

diff --git a/HISTORY.md b/HISTORY.md
index aaccbb0..58de623 100644
--- a/HISTORY.md
+++ b/HISTORY.md
@@ -9,6 +9,7 @@
 ### Bugfixes
 
 - Fixed a crash when exporting an empty string with the ReST format (#368)
+- Error cells from imported .xls files contain now the error string (#202)
 
 ## 0.14.0 (2019-10-19)
 
diff --git a/src/tablib/formats/_xls.py b/src/tablib/formats/_xls.py
index 3fc94af..0533476 100644
--- a/src/tablib/formats/_xls.py
+++ b/src/tablib/formats/_xls.py
@@ -76,7 +76,10 @@ def import_set(dset, in_stream, headers=True):
         if (i == 0) and (headers):
             dset.headers = sheet.row_values(0)
         else:
-            dset.append(sheet.row_values(i))
+            dset.append([
+                val if typ != xlrd.XL_CELL_ERROR else xlrd.error_text_from_code[val]
+                for val, typ in zip(sheet.row_values(i), sheet.row_types(i))
+            ])
 
 
 def import_book(dbook, in_stream, headers=True):
diff --git a/tests/files/errors.xls b/tests/files/errors.xls
new file mode 100644
index 0000000000000000000000000000000000000000..08693375ea12888993c6f70b8c7597d8e8d15794
GIT binary patch
literal 5632
zcmeHLU1(fI6h8OvPjAwnY|^A^f8sT@o6VoJ2DBn=H~rIsZ9}XmrJ#2=H*xL$lyuWt
zP`XX+L&1X52VV+O2#8etLDUBY(+9x^MJWVYK}h>n^kFMiDYRX`Z|=Rx?V3$DT9ry>
zIWuR@oHJ+U%$YOi?)Tr-Pn`Z}=|w3kk4r#qmTDyIqietouOAW7fs=N#R4RF7pmXIe
zvcL^@tvtgZ4lxAana=}|R^zsJ(U#_xItvT*Ne;ifn6gE(XoutwPQ3cE1Rmp8SB<1U
zUR1+>JDxfHQx>?LtvK$tKjSy!cmPm-{~7Cf|7!vB0Sf>N0gC{206YOw4`=`^1~dYi
z080S(0PY3c2UrSN23QVQ0cZx?58yf;a9iRJ0$=6!TiteQXLHy;VSyS0rx*^&=L^4K
z$XBudB0S>5;92Dtmq6Pi`{)~+taQ@+x5`Gf21}!Fj=@RS0n_9ZsEgzSMg2n6>y-C_
zDK6$(G|Wh@YH_vp6RK`e^&-@oe5Pc6P?R<B%df#-I{b!v2|`*A;$KQENlV!%`0mtb
zb8vx-@_l{6&e}s(dP3csQTerGLW@_%Owze9MZa1doz|L;zY?9ZPRIZMrH{g<Cb#Dn
z&(0l=9g5}jwhHvl3iS30^qOhu0#8uunM|w4%#%9qT0u%|BZYu!>kDDkHWY%YU0j%_
z+Qvet^n&u~V~ED*Ajk}Tt38}br|mVb<JPv}cUCh1BD+p#@{{al>@N1HjfRVi(Y)OB
z5PF**x@CTmvHeTTX~+=`Z-Fs}yCp_kny)bW<25kwDj10N#_xC#uFEr=OaZpkpE_VB
z4w<XLlC?6nIh@^_%^l2wJDba!)&VP(wi0REj1A<5GQ(-B4KGL8$o{;QPbJ$*c!Uhu
zjk);<C5<Kk5V!ErVbS?gIT0;#OGI0M{q<wti~;p(uMtdl{oO0)wkLWwJg?{lP`t8=
zX8>ay!N}QQyOQWo^@!qal$G?N=%#4Pu2ja}Z|<-Unti#9mF?Vttyt-n=$eP4Jxvnh
zgEli;uQh0!(cYtK`qxw)#ZtOexxAs(hsS^)?K(fwb-uWv&4-_s7ChAb*c^W&(y4W7
zp1jv;I2s*XuPO_Q9G}l8TvVS@XBN^tPG1L>SUlAaAU(KHKh=kEG*4c1rUQ-xxd*S)
zH55*70&krdps_<qk9I>P7EUT5;`B5e5*eJHu8z#>1&~}YO$j53AP?8WH_Q9@oF~m4
z%q_xUWFFiS8JKT)47x0D4<jSdK<8MyZ!91^wV0WL)N|_~1%H-a3KgG(5S*t-XHi?h
zmg*!>T#bI61dB06uPHh<K*v{fsa0|`13$FZ>kaa$`ofO;xK?mJE`tjA3j_vn!v?3^
z6U)<>3{Z$J|2K+;EG&RII9<e`1K_<?wQOb1wmp>`%I(h$<jpNFC+)QP*rQ#$tb~p5
z0&9tEObrY;&#ogg>l!9>l+Iex5x5B>@I1>4k7nZ7+rM1f`C{bEn}&3>etdyP^a_9r
z=QtP6{5B4t8J_|$&wUQSFw+lUzPk@V4VP~?xhwH*KpY^4*FVC(!r4cO17hu^cjWju
zqow?I>60II;6VRWDy^S$>f_&zpL%Fn<m6jef9JLLH-i=`qwN8WF`SbV=ysIlS}V~p
zJX!AK_n*>V=0lT;d<Z?;^V?_f^*JPaC854U_xiZ1LmRML?iV*o`EYeoB44R&mp0o3
z%^95WPORajYQ-n{sGs}c{b89Lm;o^{w#^;apf@e{4BRsh)k43c*{Nsdc1By!@LP=~
zBu~pQ_Mbu4%WwZ}_!nS6oe3#<WWza7Dwmk6pW&?9{jUV?d+wAW*$fRywFX=L12Mbx
Zto;l7k3x6e)qf|N+1lU#{rz9-{{snlLz(~p

literal 0
HcmV?d00001

diff --git a/tests/test_tablib.py b/tests/test_tablib.py
index 00b8214..695d13d 100755
--- a/tests/test_tablib.py
+++ b/tests/test_tablib.py
@@ -6,6 +6,8 @@ import doctest
 import json
 import pickle
 import unittest
+from collections import OrderedDict
+from pathlib import Path
 from uuid import uuid4
 
 import tablib
@@ -887,17 +889,34 @@ class TSVTests(BaseTestCase):
         self.assertEqual(tsv, self.founders.tsv)
 
 
+class XLSTests(BaseTestCase):
+    def test_xls_format_detect(self):
+        """Test the XLS format detection."""
+        in_stream = self.founders.xls
+        self.assertEqual(detect_format(in_stream), 'xls')
+
+    def test_xls_import_with_errors(self):
+        """Errors from imported files are kept as errors."""
+        xls_source = Path(__file__).parent / 'files' / 'errors.xls'
+        with xls_source.open('rb') as fh:
+            data = tablib.Dataset().load(fh.read())
+        self.assertEqual(
+            data.dict[0],
+            OrderedDict([
+                ('div by 0', '#DIV/0!'),
+                ('name unknown', '#NAME?'),
+                ('not available (formula)', '#N/A'),
+                ('not available (static)', '#N/A')
+            ])
+        )
+
+
 class XLSXTests(BaseTestCase):
     def test_xlsx_format_detect(self):
         """Test the XLSX format detection."""
         in_stream = self.founders.xlsx
         self.assertEqual(detect_format(in_stream), 'xlsx')
 
-    def test_xls_format_detect(self):
-        """Test the XLS format detection."""
-        in_stream = self.founders.xls
-        self.assertEqual(detect_format(in_stream), 'xls')
-
     def test_xlsx_import_set(self):
         date_time = datetime.datetime(2019, 10, 4, 12, 30, 8)
         data.append(('string', '004', 42, 21.55, date_time))