/******************************************************************************
 *
 * Project:  PDS Driver; Planetary Data System Format
 * Purpose:  Implementation of NASAKeywordHandler - a class to read
 *           keyword data from PDS, ISIS2 and ISIS3 data products.
 * Author:   Frank Warmerdam <warmerdam@pobox.com
 *
 ******************************************************************************
 * Copyright (c) 2006, Frank Warmerdam <warmerdam@pobox.com>
 * Copyright (c) 2008-2010, Even Rouault <even dot rouault at spatialys.com>
 * Copyright (c) 2017 Hobu Inc
 * Copyright (c) 2017, Dmitry Baryshnikov <polimax@mail.ru>
 * Copyright (c) 2017, NextGIS <info@nextgis.com>
 *
 * SPDX-License-Identifier: MIT
 ****************************************************************************
 * Object Description Language (ODL) is used to encode data labels for PDS
 * and other NASA data systems. Refer to Chapter 12 of "PDS Standards
 * Reference" at http://pds.jpl.nasa.gov/tools/standards-reference.shtml for
 * further details about ODL.
 *
 * This is also known as PVL (Parameter Value Language) which is written
 * about at http://www.orrery.us/node/44 where it notes:
 *
 * The PVL syntax that the PDS uses is specified by the Consultative Committee
 * for Space Data Systems in their Blue Book publication: "Parameter Value
 * Language Specification (CCSD0006 and CCSD0008)", June 2000
 * [CCSDS 641.0-B-2], and Green Book publication: "Parameter Value Language -
 * A Tutorial", June 2000 [CCSDS 641.0-G-2]. PVL has also been accepted by the
 * International Standards Organization (ISO), as a Final Draft International
 * Standard (ISO 14961:2002) keyword value type language for naming and
 * expressing data values.
 * --
 * also of interest, on PDS ODL:
 *  http://pds.jpl.nasa.gov/documents/sr/Chapter12.pdf
 *
 ****************************************************************************/

#include "nasakeywordhandler.h"
#include "ogrlibjsonutils.h"
#include <vector>

//! @cond Doxygen_Suppress

/************************************************************************/
/* ==================================================================== */
/*                          NASAKeywordHandler                          */
/* ==================================================================== */
/************************************************************************/

/************************************************************************/
/*                         NASAKeywordHandler()                         */
/************************************************************************/

NASAKeywordHandler::NASAKeywordHandler()
    : pszHeaderNext(nullptr), m_bStripSurroundingQuotes(false)
{
    oJSon.Deinit();
}

/************************************************************************/
/*                        ~NASAKeywordHandler()                         */
/************************************************************************/

NASAKeywordHandler::~NASAKeywordHandler()

{
}

/************************************************************************/
/*                               Ingest()                               */
/************************************************************************/

bool NASAKeywordHandler::Ingest(VSILFILE *fp, int nOffset)

{
    /* -------------------------------------------------------------------- */
    /*      Read in buffer till we find END all on its own line.            */
    /* -------------------------------------------------------------------- */
    if (VSIFSeekL(fp, nOffset, SEEK_SET) != 0)
        return false;

    std::string osHeaderText;
    for (; true;)
    {
        char szChunk[513];

        int nBytesRead = static_cast<int>(VSIFReadL(szChunk, 1, 512, fp));

        szChunk[nBytesRead] = '\0';
        osHeaderText += szChunk;

        if (nBytesRead < 512)
            break;

        const char *pszCheck = nullptr;
        if (osHeaderText.size() > 520)
            pszCheck = osHeaderText.c_str() + (osHeaderText.size() - 520);
        else
            pszCheck = szChunk;

        if (strstr(pszCheck, "\r\nEND\r\n") != nullptr ||
            strstr(pszCheck, "\nEND\n") != nullptr ||
            strstr(pszCheck, "\r\nEnd\r\n") != nullptr ||
            strstr(pszCheck, "\nEnd\n") != nullptr)
            break;
    }

    return Parse(osHeaderText.c_str());
}

/************************************************************************/
/*                               Parse()                                */
/************************************************************************/

bool NASAKeywordHandler::Parse(const char *pszStr)

{
    pszHeaderNext = pszStr;

    /* -------------------------------------------------------------------- */
    /*      Process name/value pairs, keeping track of a "path stack".      */
    /* -------------------------------------------------------------------- */
    oJSon = CPLJSONObject();
    return ReadGroup("", oJSon, 0);
}

/************************************************************************/
/*                             ReadGroup()                              */
/************************************************************************/

bool NASAKeywordHandler::ReadGroup(const std::string &osPathPrefix,
                                   CPLJSONObject &oCur, int nRecLevel)

{
    if (osPathPrefix.size() > 256)
    {
        CPLError(CE_Failure, CPLE_NotSupported, "Too big prefix for GROUP");
        return false;
    }
    if (nRecLevel == 100)
        return false;
    for (; true;)
    {
        CPLString osName, osValue;
        if (!ReadPair(osName, osValue, oCur))
            return false;

        if (EQUAL(osName, "OBJECT") || EQUAL(osName, "GROUP"))
        {
            CPLJSONObject oNewGroup;
            oNewGroup.Add("_type",
                          EQUAL(osName, "OBJECT") ? "object" : "group");
            if (!ReadGroup((osPathPrefix + osValue + ".").c_str(), oNewGroup,
                           nRecLevel + 1))
            {
                return false;
            }
            CPLJSONObject oName = oNewGroup["Name"];
            if (oName.GetType() == CPLJSONObject::Type::String)
            {
                oCur.Add(osValue + "_" + oName.ToString(), oNewGroup);
                oNewGroup.Add("_container_name", osValue);
            }
            else if (oCur[osValue].IsValid())
            {
                int nIter = 2;
                while (oCur[osValue + CPLSPrintf("_%d", nIter)].IsValid())
                {
                    nIter++;
                }
                oCur.Add(osValue + CPLSPrintf("_%d", nIter), oNewGroup);
                oNewGroup.Add("_container_name", osValue);
            }
            else
            {
                oCur.Add(osValue, oNewGroup);
            }
        }
        else if (EQUAL(osName, "END") || EQUAL(osName, "END_GROUP") ||
                 EQUAL(osName, "END_OBJECT"))
        {
            return true;
        }
        else
        {
            osName = osPathPrefix + osName;
            aosKeywordList.AddNameValue(osName, osValue);
        }
    }
}

/************************************************************************/
/*                        StripQuotesIfNeeded()                         */
/************************************************************************/

static CPLString StripQuotesIfNeeded(const CPLString &osWord,
                                     bool bQuotesAlreadyRemoved)
{
    if (bQuotesAlreadyRemoved || osWord.size() < 2 || osWord[0] != '"')
        return osWord;
    return osWord.substr(1, osWord.size() - 2);
}

/************************************************************************/
/*                              ReadPair()                              */
/*                                                                      */
/*      Read a name/value pair from the input stream.  Strip off        */
/*      white space, ignore comments, split on '='.                     */
/*      Returns TRUE on success.                                        */
/************************************************************************/

bool NASAKeywordHandler::ReadPair(CPLString &osName, CPLString &osValue,
                                  CPLJSONObject &oCur)

{
    osName = "";
    osValue = "";

    if (!ReadWord(osName))
        return false;

    SkipWhite();

    if (EQUAL(osName, "END"))
        return true;

    if (*pszHeaderNext != '=')
    {
        // ISIS3 does not have anything after the end group/object keyword.
        if (EQUAL(osName, "End_Group") || EQUAL(osName, "End_Object"))
            return true;

        return false;
    }

    pszHeaderNext++;

    SkipWhite();

    osValue = "";
    bool bIsString = true;

    // Handle value lists like:
    // Name   = (Red, Red) or  {Red, Red} or even ({Red, Red}, {Red, Red})
    CPLJSONArray oArray;
    if (*pszHeaderNext == '(' || *pszHeaderNext == '{')
    {
        std::vector<char> oStackArrayBeginChar;
        CPLString osWord;

        oStackArrayBeginChar.push_back(*pszHeaderNext);
        osValue += *pszHeaderNext;
        pszHeaderNext++;

        while (ReadWord(osWord, false, true, &bIsString))
        {
            if (*pszHeaderNext == '(' || *pszHeaderNext == '{')
            {
                oStackArrayBeginChar.push_back(*pszHeaderNext);
                osValue += *pszHeaderNext;
                pszHeaderNext++;
            }

            // TODO: we could probably do better with nested json arrays
            // instead of flattening when there are (( )) or ({ }) constructs
            if (bIsString)
            {
                if (!(osWord.empty() &&
                      (*pszHeaderNext == '(' || *pszHeaderNext == '{' ||
                       *pszHeaderNext == ')' || *pszHeaderNext == '}')))
                {
                    std::string osValueInArray =
                        StripQuotesIfNeeded(osWord, false);
                    if (!osValueInArray.empty() && osValueInArray == osWord &&
                        osValueInArray.back() == '>')
                    {
                        const auto nPosLeftBracket = osValueInArray.rfind('<');
                        if (nPosLeftBracket != std::string::npos)
                        {
                            const std::string osUnit = osValueInArray.substr(
                                nPosLeftBracket + 1, osValueInArray.size() - 1 -
                                                         (nPosLeftBracket + 1));
                            osValueInArray.resize(nPosLeftBracket);
                            while (!osValueInArray.empty() &&
                                   osValueInArray.back() == ' ')
                                osValueInArray.pop_back();

                            CPLJSONObject newObject;
                            if (CPLGetValueType(osValueInArray.c_str()) ==
                                CPL_VALUE_STRING)
                            {
                                newObject.Add("value", osValueInArray);
                            }
                            else if (CPLGetValueType(osValueInArray.c_str()) ==
                                     CPL_VALUE_INTEGER)
                            {
                                newObject.Add("value",
                                              atoi(osValueInArray.c_str()));
                            }
                            else
                            {
                                newObject.Add("value",
                                              CPLAtof(osValueInArray.c_str()));
                            }
                            newObject.Add("unit", osUnit);
                            oArray.Add(newObject);
                        }
                        else
                        {
                            oArray.Add(osValueInArray);
                        }
                    }
                    else
                    {
                        oArray.Add(osValueInArray);
                    }
                }
            }
            else if (CPLGetValueType(osWord) == CPL_VALUE_INTEGER)
            {
                oArray.Add(atoi(osWord));
            }
            else
            {
                oArray.Add(CPLAtof(osWord));
            }

            osValue += osWord;
            while (isspace(static_cast<unsigned char>(*pszHeaderNext)))
            {
                pszHeaderNext++;
            }

            if (*pszHeaderNext == ')')
            {
                osValue += *pszHeaderNext;
                if (oStackArrayBeginChar.empty() ||
                    oStackArrayBeginChar.back() != '(')
                {
                    CPLDebug("PDS", "Unpaired ( ) for %s", osName.c_str());
                    return false;
                }
                oStackArrayBeginChar.pop_back();
                pszHeaderNext++;
                if (oStackArrayBeginChar.empty())
                    break;
            }
            else if (*pszHeaderNext == '}')
            {
                osValue += *pszHeaderNext;
                if (oStackArrayBeginChar.empty() ||
                    oStackArrayBeginChar.back() != '{')
                {
                    CPLDebug("PDS", "Unpaired { } for %s", osName.c_str());
                    return false;
                }
                oStackArrayBeginChar.pop_back();
                pszHeaderNext++;
                if (oStackArrayBeginChar.empty())
                    break;
            }
            else if (*pszHeaderNext == ',')
            {
                osValue += *pszHeaderNext;
                pszHeaderNext++;
                // Do not use SkipWhite() here to avoid being confuse by
                // constructs like
                // FOO = (#123456,
                //        #123456)
                // where we could confuse the second line with a comment.
                while (isspace(static_cast<unsigned char>(*pszHeaderNext)))
                {
                    pszHeaderNext++;
                }
            }
            SkipWhite();
        }
    }

    else  // Handle more normal "single word" values.
    {
        if (!ReadWord(osValue, m_bStripSurroundingQuotes, false, &bIsString))
            return false;
    }

    SkipWhite();

    const auto AddToCur = [&oCur, &osName](const CPLJSONObject &o)
    {
        auto oExistingObjForName = oCur[osName];
        if (oExistingObjForName.IsValid())
        {
            if (oExistingObjForName["values"].GetType() ==
                CPLJSONObject::Type::Array)
            {
                oExistingObjForName["values"].ToArray().Add(o);
            }
            else
            {
                CPLJSONArray ar;
                ar.Add(oExistingObjForName);
                ar.Add(o);
                CPLJSONObject oObj;
                oObj.Add("values", ar);
                oCur.Delete(osName);
                oCur[osName] = oObj;
            }
        }
        else
        {
            oCur.Add(osName, o);
        }
    };

    // No units keyword?
    if (*pszHeaderNext != '<')
    {
        if (!EQUAL(osName, "OBJECT") && !EQUAL(osName, "GROUP"))
        {
            if (oArray.Size() > 0)
            {
                AddToCur(oArray);
            }
            else
            {
                CPLJSONObject oObj;
                if (bIsString)
                {
                    oObj = CPLJSONObject(StripQuotesIfNeeded(
                        osValue, m_bStripSurroundingQuotes));
                }
                else if (CPLGetValueType(osValue) == CPL_VALUE_INTEGER)
                {
                    oObj = CPLJSONObject(atoi(osValue));
                }
                else
                {
                    oObj = CPLJSONObject(CPLAtof(osValue));
                }
                AddToCur(oObj);
            }
        }
        return true;
    }

    CPLString osValueNoUnit(osValue);
    // Append units keyword.  For lines that like like this:
    //  MAP_RESOLUTION               = 4.0 <PIXEL/DEGREE>

    osValue += " ";

    CPLString osWord;
    CPLString osUnit;
    while (ReadWord(osWord))
    {
        SkipWhite();

        osValue += osWord;
        osUnit = osWord;
        if (osWord.back() == '>')
            break;
    }

    if (osUnit[0] == '<')
        osUnit = osUnit.substr(1);
    if (!osUnit.empty() && osUnit.back() == '>')
        osUnit = osUnit.substr(0, osUnit.size() - 1);

    CPLJSONObject newObject;

    if (oArray.Size() > 0)
    {
        newObject.Add("value", oArray);
    }
    else
    {
        if (bIsString)
        {
            newObject.Add("value", osValueNoUnit);
        }
        else if (CPLGetValueType(osValueNoUnit) == CPL_VALUE_INTEGER)
        {
            newObject.Add("value", atoi(osValueNoUnit));
        }
        else
        {
            newObject.Add("value", CPLAtof(osValueNoUnit));
        }
    }
    newObject.Add("unit", osUnit);
    AddToCur(newObject);

    return true;
}

/************************************************************************/
/*                              ReadWord()                              */
/*  Returns TRUE on success                                             */
/************************************************************************/

bool NASAKeywordHandler::ReadWord(CPLString &osWord,
                                  bool bStripSurroundingQuotes, bool bParseList,
                                  bool *pbIsString)

{
    if (pbIsString)
        *pbIsString = false;
    osWord = "";

    SkipWhite();

    if (!(*pszHeaderNext != '\0' && *pszHeaderNext != '=' &&
          !isspace(static_cast<unsigned char>(*pszHeaderNext))))
        return false;

    /* Extract a text string delimited by '\"' */
    /* Convert newlines (CR or LF) within quotes. While text strings
       support them as per ODL, the keyword list doesn't want them */
    if (*pszHeaderNext == '"')
    {
        if (pbIsString)
            *pbIsString = true;
        if (!bStripSurroundingQuotes)
            osWord += *(pszHeaderNext);
        pszHeaderNext++;
        while (*pszHeaderNext != '"')
        {
            if (*pszHeaderNext == '\0')
                return false;
            if (*pszHeaderNext == '\n')
            {
                osWord += "\\n";
                pszHeaderNext++;
                continue;
            }
            if (*pszHeaderNext == '\r')
            {
                osWord += "\\r";
                pszHeaderNext++;
                continue;
            }
            osWord += *(pszHeaderNext++);
        }
        if (!bStripSurroundingQuotes)
            osWord += *(pszHeaderNext);
        pszHeaderNext++;

        return true;
    }

    /* Extract a symbol string */
    /* These are expected to not have
       '\'' (delimiters),
       format effectors (should fit on a single line) or
       control characters.
    */
    if (*pszHeaderNext == '\'')
    {
        if (pbIsString)
            *pbIsString = true;
        if (!bStripSurroundingQuotes)
            osWord += *(pszHeaderNext);
        pszHeaderNext++;
        while (*pszHeaderNext != '\'')
        {
            if (*pszHeaderNext == '\0')
                return false;

            osWord += *(pszHeaderNext++);
        }
        if (!bStripSurroundingQuotes)
            osWord += *(pszHeaderNext);
        pszHeaderNext++;
        return true;
    }

    /*
     * Extract normal text.  Terminated by '=' or whitespace.
     *
     * A special exception is that a line may terminate with a '-'
     * which is taken as a line extender, and we suck up white space to new
     * text.
     */
    while (
        *pszHeaderNext != '\0' && *pszHeaderNext != '=' &&
        ((bParseList && *pszHeaderNext != ',' && *pszHeaderNext != '(' &&
          *pszHeaderNext != ')' && *pszHeaderNext != '{' &&
          *pszHeaderNext != '}') ||
         (!bParseList && !isspace(static_cast<unsigned char>(*pszHeaderNext)))))
    {
        osWord += *pszHeaderNext;
        pszHeaderNext++;

        if (*pszHeaderNext == '-' &&
            (pszHeaderNext[1] == 10 || pszHeaderNext[1] == 13))
        {
            pszHeaderNext += 2;
            SkipWhite();
        }
    }

    if (pbIsString)
        *pbIsString = CPLGetValueType(osWord) == CPL_VALUE_STRING;

    return true;
}

/************************************************************************/
/*                             SkipWhite()                              */
/*  Skip white spaces and C style comments                              */
/************************************************************************/

void NASAKeywordHandler::SkipWhite()

{
    for (; true;)
    {
        // Skip C style comments
        if (*pszHeaderNext == '/' && pszHeaderNext[1] == '*')
        {
            pszHeaderNext += 2;

            while (*pszHeaderNext != '\0' &&
                   (*pszHeaderNext != '*' || pszHeaderNext[1] != '/'))
            {
                pszHeaderNext++;
            }
            if (*pszHeaderNext == '\0')
                return;

            pszHeaderNext += 2;

            // consume till end of line.
            // reduce sensibility to a label error
            while (*pszHeaderNext != '\0' && *pszHeaderNext != 10 &&
                   *pszHeaderNext != 13)
            {
                pszHeaderNext++;
            }
            continue;
        }

        // Skip # style comments
        if ((*pszHeaderNext == 10 || *pszHeaderNext == 13 ||
             *pszHeaderNext == ' ' || *pszHeaderNext == '\t') &&
            pszHeaderNext[1] == '#')
        {
            pszHeaderNext += 2;

            // consume till end of line.
            while (*pszHeaderNext != '\0' && *pszHeaderNext != 10 &&
                   *pszHeaderNext != 13)
            {
                pszHeaderNext++;
            }
            continue;
        }

        // Skip white space (newline, space, tab, etc )
        if (isspace(static_cast<unsigned char>(*pszHeaderNext)))
        {
            pszHeaderNext++;
            continue;
        }

        // not white space, return.
        return;
    }
}

/************************************************************************/
/*                             GetKeyword()                             */
/************************************************************************/

const char *NASAKeywordHandler::GetKeyword(const char *pszPath,
                                           const char *pszDefault)

{
    return aosKeywordList.FetchNameValueDef(pszPath, pszDefault);
}

/************************************************************************/
/*                             GetKeywordList()                         */
/************************************************************************/

char **NASAKeywordHandler::GetKeywordList()
{
    return aosKeywordList.List();
}

/************************************************************************/
/*                               StealJSon()                            */
/************************************************************************/

CPLJSONObject NASAKeywordHandler::GetJsonObject() const
{
    return oJSon;
}

//! @endcond