You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@tika.apache.org by ta...@apache.org on 2016/02/19 21:25:29 UTC
[50/52] [partial] tika git commit: move test files to parser-modules
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testAccess_V1997.mdb
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testAccess_V1997.mdb b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testAccess_V1997.mdb
new file mode 100644
index 0000000..2aa1060
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testAccess_V1997.mdb differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testBinControlWord.rtf
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testBinControlWord.rtf b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testBinControlWord.rtf
new file mode 100644
index 0000000..9efc7cb
--- /dev/null
+++ b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testBinControlWord.rtf
@@ -0,0 +1,2 @@
+{\rtf1\ansi\ansicpg1252\uc1 \deff0\deflang1033\deflangfe1033{\fonttbl{\f0\froman\fcharset0\fprq2{\*\panose 02020603050405020304}Times New Roman;}{\f1\fswiss\fcharset0\fprq2{\*\panose 020b0604020202020204}Arial;}{\f3\froman\fcharset2\fprq2{\*\panose 05050102010706020507}Symbol;}{\f8\froman\fcharset0\fprq2{\*\panose 00000000000000000000}Tms Rmn;}{\f15\fswiss\fcharset0\fprq3{\*\panose 020b0604030504040204}Tahoma;}{\f16\froman\fcharset0\fprq2{\*\panose 00000000000000000000}Book Antiqua;}{\f18\froman\fcharset238\fprq2 Times New Roman CE;}{\f19\froman\fcharset204\fprq2 Times New Roman Cyr;}{\f21\froman\fcharset161\fprq2 Times New Roman Greek;}{\f22\froman\fcharset162\fprq2 Times New Roman Tur;}{\f23\froman\fcharset186\fprq2 Times New Roman Baltic;}{\f24\fswiss\fcharset238\fprq2 Arial CE;}{\f25\fswiss\fcharset204\fprq2 Arial Cyr;}{\f27\fswiss\fcharset161\fprq2 Arial Greek;}{\f28\fswiss\fcharset162\fprq2 Arial Tur;}{\f29\fswiss\fcharset186\fprq2 Arial Baltic;}}{\colortbl;\red0\green0\blue0;
\red0\green0\blue255;\red0\green255\blue255;\red0\green255\blue0;\red255\green0\blue255;\red255\green0\blue0;\red255\green255\blue0;\red255\green255\blue255;\red0\green0\blue128;\red0\green128\blue128;\red0\green128\blue0;\red128\green0\blue128;\red128\green0\blue0;\red128\green128\blue0;\red128\green128\blue128;\red192\green192\blue192;}{\stylesheet{\widctlpar\adjustright \f1\cgrid \snext0 Normal;}{\s1\qc\keepn\widctlpar\adjustright \f15\fs96\cgrid \sbasedon0 \snext0 heading 1;}{\s3\li360\widctlpar\adjustright \b\f8\cgrid \sbasedon0 \snext15 heading 3;}{\*\cs10 \additive Default Paragraph Font;}{\s15\li720\widctlpar\adjustright \f1\cgrid \sbasedon0 \snext15 Normal Indent;}{\s16\widctlpar\tqc\tx4320\tqr\tx8640\adjustright \f1\cgrid \sbasedon0 \snext16 header;}{\s17\widctlpar\tqc\tx4320\tqr\tx8640\adjustright \f1\fs16\cgrid \sbasedon0 \snext17 footer;}{\s18\li1440\ri1440\sb120\sa120\sl360\slmult1\widctlpar\adjustright \f1\cgrid \sbasedon0 \snext18 dsbody;}{\s19\qc\widctlpar\adjustrig
ht \f1\cgrid \sbasedon0 \snext19 dmstitle;}{\s20\fi-720\li720\widctlpar\adjustright \b\i\caps\f16\cgrid \sbasedon0 \snext20 dms;}{\s21\fi-360\li360\widctlpar{\*\pn \pnlvlbody\ilvl11\ls2047\pnrnot0\pnf3\pnstart1\pnindent360\pnhang{\pntxtb \'88}}\ls2047\ilvl11\adjustright \f1\cgrid \sbasedon0 \snext21 checklist;}{\s22\fi-2880\li2880\widctlpar\tx2880\adjustright \f1\cgrid \sbasedon0 \snext22 Indent;}{\s23\fi-288\li3150\widctlpar\tx2880\adjustright \cgrid \sbasedon22 \snext23 indent2;}{\s24\fi-360\li3528\widctlpar\tx2880\adjustright \cgrid \sbasedon22 \snext24 indent3;}{\s25\widctlpar\adjustright \b\i\caps\f16\cgrid \sbasedon0 \snext25 ds;}}{\info{\title }{\author Nicole Mendez}{\operator Nicole Mendez}{\creatim\yr2000\mo1\dy25\hr14\min4}{\revtim\yr2000\mo1\dy25\hr14\min4}{\printim\yr1999\mo1\dy27\hr15\min44}{\version3}{\edmins1}{\nofpages1}{\nofwords5}{\nofchars29}{\*\company ect}{\nofcharsws35}{\vern89}}\paperw15840\paperh12240\margl1440\margr1440 \widowctrl\ftnbj\aenddoc\noextrasprl
\prcolbl\cvmme\sprsspbf\brkfrm\swpbdr\lytprtmet\hyphcaps0\fracwidth\viewkind1\viewscale150\pgbrdrhead\pgbrdrfoot \fet0\sectd \lndscpsxn\psz1\linex0\endnhere\sectdefaultcl {\*\pnseclvl1\pnucrm\pnstart1\pnindent720\pnhang{\pntxta .}}{\*\pnseclvl2\pnucltr\pnstart1\pnindent720\pnhang{\pntxta .}}{\*\pnseclvl3\pndec\pnstart1\pnindent720\pnhang{\pntxta .}}{\*\pnseclvl4\pnlcltr\pnstart1\pnindent720\pnhang{\pntxta )}}{\*\pnseclvl5\pndec\pnstart1\pnindent720\pnhang{\pntxtb (}{\pntxta )}}{\*\pnseclvl6\pnlcltr\pnstart1\pnindent720\pnhang{\pntxtb (}{\pntxta )}}{\*\pnseclvl7\pnlcrm\pnstart1\pnindent720\pnhang{\pntxtb (}{\pntxta )}}{\*\pnseclvl8\pnlcltr\pnstart1\pnindent720\pnhang{\pntxtb (}{\pntxta )}}{\*\pnseclvl9\pnlcrm\pnstart1\pnindent720\pnhang{\pntxtb (}{\pntxta )}}\pard\plain \widctlpar\adjustright \f1\cgrid {\pard\plain \widctlpar\adjustright \f1\cgrid {\object\objemb\objw3600\objh3600\objscalex55\objscaley56{\*\objclass PBrush}{\result {\fs20 {\pict\wmetafile8\picw6350\pich6350\picwgoal3
600\pichgoal3600 \picscalex55\picscaley56 \bin10 ����}�����}}}{\tab \tab \tab
+\par }}
\ No newline at end of file
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testChm.chm
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testChm.chm b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testChm.chm
new file mode 100644
index 0000000..d7067d7
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testChm.chm differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testChm2.chm
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testChm2.chm b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testChm2.chm
new file mode 100644
index 0000000..90a6301
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testChm2.chm differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testChm3.chm
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testChm3.chm b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testChm3.chm
new file mode 100644
index 0000000..a048c46
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testChm3.chm differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testControlCharacters.doc
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testControlCharacters.doc b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testControlCharacters.doc
new file mode 100644
index 0000000..aacb46f
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testControlCharacters.doc differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testDOCX_Thumbnail.docx
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testDOCX_Thumbnail.docx b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testDOCX_Thumbnail.docx
new file mode 100644
index 0000000..85f19c7
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testDOCX_Thumbnail.docx differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testDocumentLink.doc
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testDocumentLink.doc b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testDocumentLink.doc
new file mode 100644
index 0000000..6b937c6
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testDocumentLink.doc differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL-charts.xls
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL-charts.xls b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL-charts.xls
new file mode 100644
index 0000000..52762a4
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL-charts.xls differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL-formats.xls
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL-formats.xls b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL-formats.xls
new file mode 100644
index 0000000..6929989
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL-formats.xls differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL-formats.xlsx
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL-formats.xlsx b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL-formats.xlsx
new file mode 100644
index 0000000..d4e7bce
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL-formats.xlsx differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL.strict.xlsx
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL.strict.xlsx b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL.strict.xlsx
new file mode 100644
index 0000000..6bf1936
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL.strict.xlsx differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL.xls
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL.xls b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL.xls
new file mode 100644
index 0000000..347d8a6
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL.xls differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL.xlsb
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL.xlsb b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL.xlsb
new file mode 100644
index 0000000..78fdf26
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL.xlsb differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL.xlsx
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL.xlsx b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL.xlsx
new file mode 100644
index 0000000..8d5169f
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL.xlsx differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL_1img.xls
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL_1img.xls b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL_1img.xls
new file mode 100644
index 0000000..2b447af
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL_1img.xls differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL_1img.xlsx
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL_1img.xlsx b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL_1img.xlsx
new file mode 100644
index 0000000..59f33e5
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL_1img.xlsx differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL_4.xls
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL_4.xls b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL_4.xls
new file mode 100644
index 0000000..02f58f7
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL_4.xls differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL_5.xls
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL_5.xls b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL_5.xls
new file mode 100644
index 0000000..ed69b3d
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL_5.xls differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL_95.xls
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL_95.xls b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL_95.xls
new file mode 100644
index 0000000..dac9ed3
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL_95.xls differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL_custom_props.xls
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL_custom_props.xls b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL_custom_props.xls
new file mode 100644
index 0000000..6284f9f
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL_custom_props.xls differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL_custom_props.xlsx
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL_custom_props.xlsx b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL_custom_props.xlsx
new file mode 100644
index 0000000..24eb0d5
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL_custom_props.xlsx differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL_embeded.xls
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL_embeded.xls b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL_embeded.xls
new file mode 100644
index 0000000..b72cc0b
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL_embeded.xls differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL_embeded.xlsx
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL_embeded.xlsx b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL_embeded.xlsx
new file mode 100644
index 0000000..e68d9ba
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL_embeded.xlsx differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL_headers_footers.xls
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL_headers_footers.xls b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL_headers_footers.xls
new file mode 100644
index 0000000..2c87e1c
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL_headers_footers.xls differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL_headers_footers.xlsx
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL_headers_footers.xlsx b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL_headers_footers.xlsx
new file mode 100644
index 0000000..e873323
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL_headers_footers.xlsx differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL_protected_passtika.xls
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL_protected_passtika.xls b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL_protected_passtika.xls
new file mode 100644
index 0000000..8d7e1ec
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL_protected_passtika.xls differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL_protected_passtika.xlsx
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL_protected_passtika.xlsx b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL_protected_passtika.xlsx
new file mode 100644
index 0000000..deea286
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL_protected_passtika.xlsx differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL_textbox.xlsx
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL_textbox.xlsx b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL_textbox.xlsx
new file mode 100644
index 0000000..d4b7b3d
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testEXCEL_textbox.xlsx differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testException1.doc
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testException1.doc b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testException1.doc
new file mode 100644
index 0000000..77446f3
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testException1.doc differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testException2.doc
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testException2.doc b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testException2.doc
new file mode 100644
index 0000000..3ad75eb
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testException2.doc differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testFontAfterBufferedText.rtf
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testFontAfterBufferedText.rtf b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testFontAfterBufferedText.rtf
new file mode 100644
index 0000000..de6dc2f
--- /dev/null
+++ b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testFontAfterBufferedText.rtf
@@ -0,0 +1,7 @@
+{\rtf1\ansi\ansicpg1252\fromtext \fbidis \deff0
+{\fonttbl
+
+{\f0\fswiss\fcharset0 Arial;} {\f1\fswiss\fcharset204 Arial;}
+}
+\par{\f1\fs20 \'d3\'e2\'e0\'e6\'e0\'e5\'ec\'fb\'e9 \'ea\'eb\'e8\'e5\'ed\'f2!\f0}\par
+}
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testFooter.ods
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testFooter.ods b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testFooter.ods
new file mode 100644
index 0000000..fddac2c
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testFooter.ods differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testFooter.odt
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testFooter.odt b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testFooter.odt
new file mode 100644
index 0000000..5dd4174
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testFooter.odt differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testMSG.msg
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testMSG.msg b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testMSG.msg
new file mode 100644
index 0000000..74419c4
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testMSG.msg differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testMSG_att_doc.msg
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testMSG_att_doc.msg b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testMSG_att_doc.msg
new file mode 100644
index 0000000..ea82d91
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testMSG_att_doc.msg differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testMSG_att_msg.msg
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testMSG_att_msg.msg b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testMSG_att_msg.msg
new file mode 100644
index 0000000..35b66e1
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testMSG_att_msg.msg differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testMSG_chinese.msg
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testMSG_chinese.msg b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testMSG_chinese.msg
new file mode 100644
index 0000000..c2b84c0
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testMSG_chinese.msg differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testMSG_forwarded.msg
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testMSG_forwarded.msg b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testMSG_forwarded.msg
new file mode 100644
index 0000000..d74d31a
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testMSG_forwarded.msg differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testMasterFooter.odp
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testMasterFooter.odp b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testMasterFooter.odp
new file mode 100644
index 0000000..4108a4d
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testMasterFooter.odp differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testNPEOpenDocument.odt
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testNPEOpenDocument.odt b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testNPEOpenDocument.odt
new file mode 100644
index 0000000..57f4369
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testNPEOpenDocument.odt differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testODFwithOOo3.odt
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testODFwithOOo3.odt b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testODFwithOOo3.odt
new file mode 100644
index 0000000..7915c6e
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testODFwithOOo3.odt differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testODT-TIKA-6000.odt
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testODT-TIKA-6000.odt b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testODT-TIKA-6000.odt
new file mode 100644
index 0000000..a6e3e63
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testODT-TIKA-6000.odt differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testOpenOffice2.odf
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testOpenOffice2.odf b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testOpenOffice2.odf
new file mode 100644
index 0000000..a814972
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testOpenOffice2.odf differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testOpenOffice2.odt
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testOpenOffice2.odt b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testOpenOffice2.odt
new file mode 100644
index 0000000..bc31925
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testOpenOffice2.odt differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPM.ppm
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPM.ppm b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPM.ppm
new file mode 100644
index 0000000..63ad5b1
--- /dev/null
+++ b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPM.ppm
@@ -0,0 +1,4 @@
+P3
+1 1
+255
+0 0 0
\ No newline at end of file
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT.potm
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT.potm b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT.potm
new file mode 100644
index 0000000..b78da7f
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT.potm differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT.ppsm
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT.ppsm b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT.ppsm
new file mode 100644
index 0000000..6ba1432
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT.ppsm differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT.ppsx
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT.ppsx b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT.ppsx
new file mode 100644
index 0000000..814ec0e
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT.ppsx differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT.ppt
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT.ppt b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT.ppt
new file mode 100644
index 0000000..b48cfaf
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT.ppt differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT.pptm
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT.pptm b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT.pptm
new file mode 100644
index 0000000..e8fe162
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT.pptm differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT.pptx
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT.pptx b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT.pptx
new file mode 100644
index 0000000..38bc6a2
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT.pptx differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT.thmx
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT.thmx b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT.thmx
new file mode 100644
index 0000000..9144088
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT.thmx differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT.xps
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT.xps b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT.xps
new file mode 100644
index 0000000..678033a
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT.xps differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPTX_Thumbnail.pptx
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPTX_Thumbnail.pptx b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPTX_Thumbnail.pptx
new file mode 100644
index 0000000..ab4c62a
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPTX_Thumbnail.pptx differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_2imgs.pptx
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_2imgs.pptx b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_2imgs.pptx
new file mode 100644
index 0000000..d459a6c
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_2imgs.pptx differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_autodate.ppt
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_autodate.ppt b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_autodate.ppt
new file mode 100644
index 0000000..4abb8ac
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_autodate.ppt differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_autodate.pptx
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_autodate.pptx b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_autodate.pptx
new file mode 100644
index 0000000..63aaf76
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_autodate.pptx differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_comment.ppt
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_comment.ppt b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_comment.ppt
new file mode 100644
index 0000000..7f092c7
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_comment.ppt differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_comment.pptx
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_comment.pptx b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_comment.pptx
new file mode 100644
index 0000000..f408073
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_comment.pptx differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_custom_props.ppt
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_custom_props.ppt b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_custom_props.ppt
new file mode 100644
index 0000000..82dd1e2
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_custom_props.ppt differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_custom_props.pptx
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_custom_props.pptx b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_custom_props.pptx
new file mode 100644
index 0000000..091fa0a
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_custom_props.pptx differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_embedded2.ppt
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_embedded2.ppt b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_embedded2.ppt
new file mode 100644
index 0000000..b0de26d
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_embedded2.ppt differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_embedded_two_slides.pptx
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_embedded_two_slides.pptx b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_embedded_two_slides.pptx
new file mode 100644
index 0000000..210a4bc
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_embedded_two_slides.pptx differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_embeded.ppt
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_embeded.ppt b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_embeded.ppt
new file mode 100644
index 0000000..44d47ee
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_embeded.ppt differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_embeded.pptx
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_embeded.pptx b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_embeded.pptx
new file mode 100644
index 0000000..d0a97cd
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_embeded.pptx differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_masterFooter.ppt
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_masterFooter.ppt b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_masterFooter.ppt
new file mode 100644
index 0000000..161d0a4
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_masterFooter.ppt differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_masterFooter.pptx
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_masterFooter.pptx b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_masterFooter.pptx
new file mode 100644
index 0000000..70da94c
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_masterFooter.pptx differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_masterText.ppt
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_masterText.ppt b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_masterText.ppt
new file mode 100644
index 0000000..cdcf4bc
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_masterText.ppt differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_masterText.pptx
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_masterText.pptx b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_masterText.pptx
new file mode 100644
index 0000000..46dbc36
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_masterText.pptx differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_masterText2.ppt
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_masterText2.ppt b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_masterText2.ppt
new file mode 100644
index 0000000..d042aac
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_masterText2.ppt differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_masterText2.pptx
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_masterText2.pptx b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_masterText2.pptx
new file mode 100644
index 0000000..25c56a9
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_masterText2.pptx differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_protected_passtika.ppt
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_protected_passtika.ppt b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_protected_passtika.ppt
new file mode 100644
index 0000000..ae67450
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_protected_passtika.ppt differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_protected_passtika.pptx
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_protected_passtika.pptx b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_protected_passtika.pptx
new file mode 100644
index 0000000..d242870
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_protected_passtika.pptx differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_various.ppt
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_various.ppt b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_various.ppt
new file mode 100644
index 0000000..6f72715
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_various.ppt differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_various.pptx
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_various.pptx b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_various.pptx
new file mode 100644
index 0000000..92c2744
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPPT_various.pptx differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPROJECT2003.mpp
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPROJECT2003.mpp b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPROJECT2003.mpp
new file mode 100644
index 0000000..69bad2c
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPROJECT2003.mpp differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPROJECT2007.mpp
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPROJECT2007.mpp b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPROJECT2007.mpp
new file mode 100644
index 0000000..964b2b0
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPROJECT2007.mpp differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPST.pst
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPST.pst b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPST.pst
new file mode 100644
index 0000000..005e215
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPST.pst differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPUBLISHER.pub
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPUBLISHER.pub b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPUBLISHER.pub
new file mode 100644
index 0000000..2397b9d
Binary files /dev/null and b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testPUBLISHER.pub differ
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testRTF-ms932.rtf
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testRTF-ms932.rtf b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testRTF-ms932.rtf
new file mode 100644
index 0000000..c5e5794
--- /dev/null
+++ b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testRTF-ms932.rtf
@@ -0,0 +1,30 @@
+{\rtf1\ansi\ansicpg932\uc2 \deff26\deflang1033\deflangfe1041{\fonttbl{\f0\froman\fcharset0\fprq2{\*\panose 02020603050405020304}Times New Roman;}
+{\f23\froman\fcharset128\fprq1{\*\panose 02020609040205080304}\'82\'6c\'82\'72 \'96\'be\'92\'a9{\*\falt MS Mincho};}{\f26\froman\fcharset0\fprq2{\*\panose 02040604050505020304}Century;}
+{\f28\froman\fcharset128\fprq1{\*\panose 02020609040205080304}@\'82\'6c\'82\'72 \'96\'be\'92\'a9;}{\f37\froman\fcharset238\fprq2 Times New Roman CE;}{\f38\froman\fcharset204\fprq2 Times New Roman Cyr;}{\f40\froman\fcharset161\fprq2 Times New Roman Greek;}
+{\f41\froman\fcharset162\fprq2 Times New Roman Tur;}{\f42\froman\fcharset177\fprq2 Times New Roman (Hebrew);}{\f43\froman\fcharset178\fprq2 Times New Roman (Arabic);}{\f44\froman\fcharset186\fprq2 Times New Roman Baltic;}
+{\f223\froman\fcharset0\fprq1 MS Mincho Western{\*\falt MS Mincho};}{\f221\froman\fcharset238\fprq1 MS Mincho CE{\*\falt MS Mincho};}{\f222\froman\fcharset204\fprq1 MS Mincho Cyr{\*\falt MS Mincho};}
+{\f224\froman\fcharset161\fprq1 MS Mincho Greek{\*\falt MS Mincho};}{\f225\froman\fcharset162\fprq1 MS Mincho Tur{\*\falt MS Mincho};}{\f228\froman\fcharset186\fprq1 MS Mincho Baltic{\*\falt MS Mincho};}{\f245\froman\fcharset238\fprq2 Century CE;}
+{\f246\froman\fcharset204\fprq2 Century Cyr;}{\f248\froman\fcharset161\fprq2 Century Greek;}{\f249\froman\fcharset162\fprq2 Century Tur;}{\f252\froman\fcharset186\fprq2 Century Baltic;}
+{\f263\froman\fcharset0\fprq1 @\'82\'6c\'82\'72 \'96\'be\'92\'a9 Western;}{\f261\froman\fcharset238\fprq1 @\'82\'6c\'82\'72 \'96\'be\'92\'a9 CE;}{\f262\froman\fcharset204\fprq1 @\'82\'6c\'82\'72 \'96\'be\'92\'a9 Cyr;}
+{\f264\froman\fcharset161\fprq1 @\'82\'6c\'82\'72 \'96\'be\'92\'a9 Greek;}{\f265\froman\fcharset162\fprq1 @\'82\'6c\'82\'72 \'96\'be\'92\'a9 Tur;}{\f268\froman\fcharset186\fprq1 @\'82\'6c\'82\'72 \'96\'be\'92\'a9 Baltic;}}{\colortbl;\red0\green0\blue0;
+\red0\green0\blue255;\red0\green255\blue255;\red0\green255\blue0;\red255\green0\blue255;\red255\green0\blue0;\red255\green255\blue0;\red255\green255\blue255;\red0\green0\blue128;\red0\green128\blue128;\red0\green128\blue0;\red128\green0\blue128;
+\red128\green0\blue0;\red128\green128\blue0;\red128\green128\blue128;\red192\green192\blue192;}{\stylesheet{\qj \li0\ri0\nowidctlpar\aspalpha\aspnum\faauto\adjustright\rin0\lin0\itap0
+\fs21\lang1033\langfe1041\kerning2\loch\f26\hich\af26\dbch\af23\cgrid\langnp1033\langfenp1041 \snext0 Normal;}{\*\cs10 \additive Default Paragraph Font;}}{\info{\title \'83\'5e\'83\'43\'83\'67\'83\'8b}{\author shinsuke}{\operator shinsuke}
+{\creatim\yr2010\mo10\dy8\hr14\min18}{\revtim\yr2010\mo10\dy10\hr6\min59}{\version4}{\edmins3}{\nofpages1}{\nofwords3}{\nofchars3}{\nofcharsws0}{\vern8249}}\paperw11906\paperh16838\margl1701\margr1701\margt1985\margb1701\gutter0
+\deftab840\ftnbj\aenddoc\hyphcaps0\formshade\horzdoc\dgmargin\dghspace180\dgvspace180\dghorigin1701\dgvorigin1985\dghshow0\dgvshow2\jcompress\lnongrid
+\viewkind1\viewscale100\splytwnine\ftnlytwnine\htmautsp\useltbaln\alntblind\lytcalctblwd\lyttblrtgr\lnbrkrule {\upr{\*\fchars
+!%),.:\'3b?]\'7d\'81\'91\'81\'8b\'81\'66\'81\'68\'81\'f1\'81\'8c\'81\'8d\'81\'8e\'81\'41\'81\'42\'81\'58\'81\'72\'81\'74\'81\'76\'81\'78\'81\'7a\'81\'6c\'81\'4a\'81\'4b\'81\'54\'81\'55\'81\'45\'81\'52\'81\'53\'81\'49\'81\'93\'81\'6a\'81\'43\'81\'44
+\'81\'46\'81\'47\'81\'48\'81\'6e\'81\'70\'a1\'a3\'a4\'a5\'de\'df\'81\'91}{\*\ud\uc0{\*\fchars
+!%),.:\'3b?]\'7d{\uc2\u162 \'81\'91\'81\'8b\'81f\'81h\'81\'f1\'81\'8c\'81\'8d\'81\'8e\'81A\'81B\'81X\'81r\'81t\'81v\'81x\'81z\'81l\'81J\'81K\'81T\'81U\'81E\'81R\'81S\'81I\'81\'93\'81j\'81C\'81D\'81F\'81G\'81H\'81n\'81p\'a1\'a3\'a4\'a5}\'de\'df\'81\'91}}}
+{\upr{\*\lchars $([\'5c\'7b\'81\'92\'5c\'81\'65\'81\'67\'81\'71\'81\'73\'81\'75\'81\'77\'81\'79\'81\'6b\'81\'90\'81\'69\'81\'6d\'81\'6f\'a2\'81\'92\'81\'8f}{\*\ud\uc0{\*\lchars
+$([\'5c\'7b{\uc2\u163 \'81\'92}{\uc1\u165 \'5c\'81e\'81g\'81q\'81s\'81u\'81w\'81y\'81k\'81\'90\'81i\'81m\'81o\'a2\'81\'92\'81\'8f}}}}\fet0\sectd \linex0\headery851\footery992\colsx425\endnhere\sectlinegrid360\sectspecifyl {\*\pnseclvl1
+\pnucrm\pnstart1\pnindent720\pnhang{\pntxta \dbch .}}{\*\pnseclvl2\pnucltr\pnstart1\pnindent720\pnhang{\pntxta \dbch .}}{\*\pnseclvl3\pndec\pnstart1\pnindent720\pnhang{\pntxta \dbch .}}{\*\pnseclvl4\pnlcltr\pnstart1\pnindent720\pnhang{\pntxta \dbch )}}
+{\*\pnseclvl5\pndec\pnstart1\pnindent720\pnhang{\pntxtb \dbch (}{\pntxta \dbch )}}{\*\pnseclvl6\pnlcltr\pnstart1\pnindent720\pnhang{\pntxtb \dbch (}{\pntxta \dbch )}}{\*\pnseclvl7\pnlcrm\pnstart1\pnindent720\pnhang{\pntxtb \dbch (}{\pntxta \dbch )}}
+{\*\pnseclvl8\pnlcltr\pnstart1\pnindent720\pnhang{\pntxtb \dbch (}{\pntxta \dbch )}}{\*\pnseclvl9\pnlcrm\pnstart1\pnindent720\pnhang{\pntxtb \dbch (}{\pntxta \dbch )}}\pard\plain \qj \li0\ri0\nowidctlpar\aspalpha\aspnum\faauto\adjustright\rin0\lin0\itap0
+\fs21\lang1033\langfe1041\kerning2\loch\af26\hich\af26\dbch\af23\cgrid\langnp1033\langfenp1041 {\hich\af26\dbch\af23\loch\f26 Hello
+\par }{\loch\af26\hich\af26\dbch\f23 \'82\'b1\'82\'f1\'82\'c9\'82\'bf\'82\'cd}{
+\par \hich\af26\dbch\af23\loch\f26 Test
+\par }{\loch\af26\hich\af26\dbch\f23 \'83\'65\'83\'58\'83\'67}{
+\par
+\par }{
+\par }}
http://git-wip-us.apache.org/repos/asf/tika/blob/38916f89/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testRTF.rtf
----------------------------------------------------------------------
diff --git a/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testRTF.rtf b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testRTF.rtf
new file mode 100644
index 0000000..487e6f4
--- /dev/null
+++ b/tika-parser-modules/tika-parser-office-module/src/test/resources/test-documents/testRTF.rtf
@@ -0,0 +1,17 @@
+{\rtf1\ansi\ansicpg1252\uc1\deff0\stshfdbch0\stshfloch0\stshfhich0\stshfbi0\deflang1036\deflangfe1036{\fonttbl{\f0\froman\fcharset0\fprq2{\*\panose 02020603050405020304}Times New Roman;}{\f37\froman\fcharset238\fprq2 Times New Roman CE;}
+{\f38\froman\fcharset204\fprq2 Times New Roman Cyr;}{\f40\froman\fcharset161\fprq2 Times New Roman Greek;}{\f41\froman\fcharset162\fprq2 Times New Roman Tur;}{\f42\froman\fcharset177\fprq2 Times New Roman (Hebrew);}
+{\f43\froman\fcharset178\fprq2 Times New Roman (Arabic);}{\f44\froman\fcharset186\fprq2 Times New Roman Baltic;}{\f45\froman\fcharset163\fprq2 Times New Roman (Vietnamese);}}{\colortbl;\red0\green0\blue0;\red0\green0\blue255;\red0\green255\blue255;
+\red0\green255\blue0;\red255\green0\blue255;\red255\green0\blue0;\red255\green255\blue0;\red255\green255\blue255;\red0\green0\blue128;\red0\green128\blue128;\red0\green128\blue0;\red128\green0\blue128;\red128\green0\blue0;\red128\green128\blue0;
+\red128\green128\blue128;\red192\green192\blue192;}{\stylesheet{\ql \li0\ri0\widctlpar\aspalpha\aspnum\faauto\adjustright\rin0\lin0\itap0 \fs24\lang1036\langfe1036\cgrid\langnp1036\langfenp1036 \snext0 Normal;}{\*\cs10 \additive \ssemihidden
+Default Paragraph Font;}{\*\ts11\tsrowd\trftsWidthB3\trpaddl108\trpaddr108\trpaddfl3\trpaddft3\trpaddfb3\trpaddfr3\trcbpat1\trcfpat1\tscellwidthfts0\tsvertalt\tsbrdrt\tsbrdrl\tsbrdrb\tsbrdrr\tsbrdrdgl\tsbrdrdgr\tsbrdrh\tsbrdrv
+\ql \li0\ri0\widctlpar\aspalpha\aspnum\faauto\adjustright\rin0\lin0\itap0 \fs20\lang1024\langfe1024\cgrid\langnp1024\langfenp1024 \snext11 \ssemihidden Normal Table;}}{\*\latentstyles\lsdstimax156\lsdlockeddef0}{\*\rsidtbl \rsid2954171\rsid10375891}
+{\*\generator Microsoft Word 11.0.6568;}{\info{\title Test d\'92indexation Word}{\author Bibliotheque}{\operator Bibliotheque}{\creatim\yr2006\mo5\dy18\hr12\min19}{\revtim\yr2006\mo5\dy18\hr12\min19}{\version2}{\edmins0}{\nofpages1}{\nofwords3}
+{\nofchars21}{\*\company Universite Laval}{\nofcharsws23}{\vern24579}}\paperw11906\paperh16838\margl1417\margr1417\margt1417\margb1417
+\deftab708\widowctrl\ftnbj\aenddoc\hyphhotz425\noxlattoyen\expshrtn\noultrlspc\dntblnsbdb\nospaceforul\formshade\horzdoc\dgmargin\dghspace180\dgvspace180\dghorigin1417\dgvorigin1417\dghshow1\dgvshow1
+\jexpand\viewkind1\viewscale100\pgbrdrhead\pgbrdrfoot\splytwnine\ftnlytwnine\htmautsp\nolnhtadjtbl\useltbaln\alntblind\lytcalctblwd\lyttblrtgr\lnbrkrule\nobrkwrptbl\snaptogridincell\allowfieldendsel\wrppunct\asianbrkrule\nojkernpunct\rsidroot2954171 \fet0
+\sectd \linex0\headery708\footery708\colsx708\endnhere\sectlinegrid360\sectdefaultcl\sftnbj {\*\pnseclvl1\pnucrm\pnstart1\pnindent720\pnhang {\pntxta .}}{\*\pnseclvl2\pnucltr\pnstart1\pnindent720\pnhang {\pntxta .}}{\*\pnseclvl3
+\pndec\pnstart1\pnindent720\pnhang {\pntxta .}}{\*\pnseclvl4\pnlcltr\pnstart1\pnindent720\pnhang {\pntxta )}}{\*\pnseclvl5\pndec\pnstart1\pnindent720\pnhang {\pntxtb (}{\pntxta )}}{\*\pnseclvl6\pnlcltr\pnstart1\pnindent720\pnhang {\pntxtb (}{\pntxta )}}
+{\*\pnseclvl7\pnlcrm\pnstart1\pnindent720\pnhang {\pntxtb (}{\pntxta )}}{\*\pnseclvl8\pnlcltr\pnstart1\pnindent720\pnhang {\pntxtb (}{\pntxta )}}{\*\pnseclvl9\pnlcrm\pnstart1\pnindent720\pnhang {\pntxtb (}{\pntxta )}}\pard\plain
+\ql \li0\ri0\widctlpar\aspalpha\aspnum\faauto\adjustright\rin0\lin0\itap0 \fs24\lang1036\langfe1036\cgrid\langnp1036\langfenp1036 {\insrsid2954171 Test d\rquote indexation Word
+\par
+\par }}
\ No newline at end of file