X-Git-Url: http://source.jalview.org/gitweb/?a=blobdiff_plain;f=src%2Fjalview%2Fio%2FIdentifyFile.java;h=4fb2516873b74e99d53dfa747ac56565da6591aa;hb=227ee1c1c69c228b67d408a0a83ae51bf5811272;hp=dc341b9ef00ba2bf31e4ac462bdb281cda2e79b1;hpb=153dd62dc91da13ae732600e6ea55ddbe15eab39;p=jalview.git
diff --git a/src/jalview/io/IdentifyFile.java b/src/jalview/io/IdentifyFile.java
index dc341b9..4fb2516 100755
--- a/src/jalview/io/IdentifyFile.java
+++ b/src/jalview/io/IdentifyFile.java
@@ -1,28 +1,30 @@
/*
- * Jalview - A Sequence Alignment Editor and Viewer (Version 2.6)
- * Copyright (C) 2010 J Procter, AM Waterhouse, G Barton, M Clamp, S Searle
- *
+ * Jalview - A Sequence Alignment Editor and Viewer (Version 2.8.2)
+ * Copyright (C) 2014 The Jalview Authors
+ *
* This file is part of Jalview.
- *
+ *
* Jalview is free software: you can redistribute it and/or
- * modify it under the terms of the GNU General Public License
- * as published by the Free Software Foundation, either version 3 of the License, or (at your option) any later version.
- *
- * Jalview is distributed in the hope that it will be useful, but
- * WITHOUT ANY WARRANTY; without even the implied warranty
- * of MERCHANTABILITY or FITNESS FOR A PARTICULAR
+ * modify it under the terms of the GNU General Public License
+ * as published by the Free Software Foundation, either version 3
+ * of the License, or (at your option) any later version.
+ *
+ * Jalview is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty
+ * of MERCHANTABILITY or FITNESS FOR A PARTICULAR
* PURPOSE. See the GNU General Public License for more details.
- *
- * You should have received a copy of the GNU General Public License along with Jalview. If not, see .
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with Jalview. If not, see .
+ * The Jalview Authors are detailed in the 'AUTHORS' file.
*/
package jalview.io;
-import java.io.*;
-import java.net.*;
+import java.io.IOException;
/**
* DOCUMENT ME!
- *
+ *
* @author $author$
* @version $Revision$
*/
@@ -30,10 +32,10 @@ public class IdentifyFile
{
/**
* Identify a datasource's file content.
- *
+ *
* @note Do not use this method for stream sources - create a FileParse object
* instead.
- *
+ *
* @param file
* DOCUMENT ME!
* @param protocol
@@ -58,7 +60,9 @@ public class IdentifyFile
emessage = e.getMessage();
}
if (parser != null)
+ {
return parser.errormessage;
+ }
return emessage;
}
@@ -71,7 +75,7 @@ public class IdentifyFile
/**
* Identify contents of source, closing it or resetting source to start
* afterwards.
- *
+ *
* @param source
* @param closeSource
* @return filetype string
@@ -132,6 +136,19 @@ public class IdentifyFile
break;
}
+ // if (data.matches("<(\"[^\"]*\"|'[^']*'|[^'\">])*>"))
+ if (data.matches("<(?i)html(\"[^\"]*\"|'[^']*'|[^'\">])*>"))
+ {
+ reply = "HTML";
+ break;
+ }
+
+ if (data.matches("<(?i)rnaml (\"[^\"]*\"|'[^']*'|[^'\">])*>"))
+ {
+ reply = "RNAML";
+
+ break;
+ }
if ((data.length() < 1) || (data.indexOf("#") == 0))
{
@@ -160,10 +177,11 @@ public class IdentifyFile
break;
}
+
else if (data.indexOf(">") > -1)
{
// FASTA, PIR file or BLC file
- boolean checkPIR = false;
+ boolean checkPIR = false, starterm = false;
if ((data.indexOf(">P1;") > -1) || (data.indexOf(">DL;") > -1))
{
// watch for PIR file attributes
@@ -180,14 +198,25 @@ public class IdentifyFile
else
{
// Is this a single line BLC file?
- source.nextLine();
+ String data1 = source.nextLine();
String data2 = source.nextLine();
- if (data2 != null && data.indexOf("*") > -1)
+ int c1;
+ if (checkPIR)
{
- if (data.indexOf("*") == data2.indexOf("*"))
+ starterm = (data1 != null && data1.indexOf("*") > -1)
+ || (data2 != null && data2.indexOf("*") > -1);
+ }
+ if (data2 != null && (c1 = data.indexOf("*")) > -1)
+ {
+ if (c1 == 0 && c1 == data2.indexOf("*"))
{
reply = "BLC";
}
+ else
+ {
+ reply = "FASTA"; // possibly a bad choice - may be recognised as
+ // PIR
+ }
// otherwise can still possibly be a PIR file
}
else
@@ -196,13 +225,49 @@ public class IdentifyFile
// TODO : AMSA File is indicated if there is annotation in the
// FASTA file - but FASTA will automatically generate this at the
// mo.
- break;
+ if (!checkPIR)
+ {
+ break;
+ }
}
}
- // TODO final check for PIR content. require
+ // final check for PIR content. require
// >P1;title\n\nterminated sequence to occur at least once.
- // the PIR/fasta ambiguity may be the use case that is needed to have
+
+ // TODO the PIR/fasta ambiguity may be the use case that is needed to
+ // have
// a 'Parse as type XXX' parameter for the applet/application.
+ if (checkPIR)
+ {
+ String dta = null;
+ if (!starterm)
+ {
+ do
+ {
+ try
+ {
+ dta = source.nextLine();
+ } catch (IOException ex)
+ {
+ }
+ ;
+ if (dta != null && dta.indexOf("*") > -1)
+ {
+ starterm = true;
+ }
+ } while (dta != null && !starterm);
+ }
+ if (starterm)
+ {
+ reply = "PIR";
+ break;
+ }
+ else
+ {
+ reply = "FASTA"; // probably a bad choice!
+ }
+ }
+ // read as a FASTA (probably)
break;
}
else if (data.indexOf("HEADER") == 0 || data.indexOf("ATOM") == 0)
@@ -210,18 +275,25 @@ public class IdentifyFile
reply = "PDB";
break;
}
+ else if (data.matches("\\s*\\d+\\s+\\d+\\s*"))
+ {
+ reply = PhylipFile.FILE_DESC;
+ break;
+ }
+
+
/*
* // TODO comment out SimpleBLAST identification for Jalview 2.4.1 else
* if (!lineswereskipped && data.indexOf("BLAST")<4) { reply =
* "SimpleBLAST"; break;
- *
+ *
* } // end comments for Jalview 2.4.1
*/
else if (!lineswereskipped && data.charAt(0) != '*'
&& data.charAt(0) != ' '
&& data.indexOf(":") < data.indexOf(",")) // &&
- // data.indexOf(",")